Skip to main content

Configuration reference

This page documents all of the configuration properties for each Druid service type.

A recommended way of organizing Druid configuration files can be seen in the conf directory in the Druid package root, shown below:

$ ls -R conf
druid

conf/druid:
_common broker coordinator historical middleManager overlord

conf/druid/_common:
common.runtime.properties log4j2.xml

conf/druid/broker:
jvm.config runtime.properties

conf/druid/coordinator:
jvm.config runtime.properties

conf/druid/historical:
jvm.config runtime.properties

conf/druid/middleManager:
jvm.config runtime.properties

conf/druid/overlord:
jvm.config runtime.properties

Each directory has a runtime.properties file containing configuration properties for the specific Druid service corresponding to the directory, such as historical.

The jvm.config files contain JVM flags such as heap sizing properties for each service.

Common properties shared by all services are placed in _common/common.runtime.properties.

Configuration interpolation

Configuration values can be interpolated from System Properties, Environment Variables, or local files. Below is an example of how this can be used:

druid.metadata.storage.type=${env:METADATA_STORAGE_TYPE}
druid.processing.tmpDir=${sys:java.io.tmpdir}
druid.segmentCache.locations=${file:UTF-8:/config/segment-cache-def.json}

Interpolation is also recursive so you can do:

druid.segmentCache.locations=${file:UTF-8:${env:SEGMENT_DEF_LOCATION}}

If the property is not set, an exception will be thrown on startup, but a default can be provided if desired. Setting a default value will not work with file interpolation as an exception will be thrown if the file does not exist.

druid.metadata.storage.type=${env:METADATA_STORAGE_TYPE:-mysql}
druid.processing.tmpDir=${sys:java.io.tmpdir:-/tmp}

If you need to set a variable that is wrapped by ${...} but do not want it to be interpolated, you can escape it by adding another $. For example:

config.name=$${value}

Common configurations

The properties under this section are common configurations that should be shared across all Druid services in a cluster.

JVM configuration best practices

There are four JVM parameters that we set on all of our services:

  • -Duser.timezone=UTC: This sets the default timezone of the JVM to UTC. We always set this and do not test with other default timezones, so local timezones might work, but they also might uncover weird and interesting bugs. To issue queries in a non-UTC timezone, see query granularities

  • -Dfile.encoding=UTF-8 This is similar to timezone, we test assuming UTF-8. Local encodings might work, but they also might result in weird and interesting bugs.

  • -Djava.io.tmpdir=<a path> Various parts of Druid use temporary files to interact with the file system. These files can become quite large. This means that systems that have small /tmp directories can cause problems for Druid. Therefore, set the JVM tmp directory to a location with ample space.

    Also consider the following when configuring the JVM tmp directory:

    • The temp directory should not be volatile tmpfs.
    • This directory should also have good read and write speed.
    • Avoid NFS mount.
    • The org.apache.druid.java.util.metrics.SysMonitor requires execute privileges on files in java.io.tmpdir. If you are using the system monitor, do not set java.io.tmpdir to noexec.
  • -Djava.util.logging.manager=org.apache.logging.log4j.jul.LogManager This allows log4j2 to handle logs for non-log4j2 components (like jetty) which use standard java logging.

Extensions

Many of Druid's external dependencies can be plugged in as modules. Extensions can be provided using the following configs:

PropertyDescriptionDefault
druid.extensions.directoryThe root extension directory where user can put extensions related files. Druid will load extensions stored under this directory.extensions (This is a relative path to Druid's working directory)
druid.extensions.hadoopDependenciesDirThe root Hadoop dependencies directory where user can put Hadoop related dependencies files. Druid will load the dependencies based on the Hadoop coordinate specified in the Hadoop index task.hadoop-dependencies (This is a relative path to Druid's working directory
druid.extensions.loadListA JSON array of extensions to load from extension directories by Druid. If it is not specified, its value will be null and Druid will load all the extensions under druid.extensions.directory. If its value is empty list [], then no extensions will be loaded at all. It is also allowed to specify absolute path of other custom extensions not stored in the common extensions directory.null
druid.extensions.searchCurrentClassloaderThis is a boolean flag that determines if Druid will search the main classloader for extensions. It defaults to true but can be turned off if you have reason to not automatically add all modules on the classpath.true
druid.extensions.useExtensionClassloaderFirstThis is a boolean flag that determines if Druid extensions should prefer loading classes from their own jars rather than jars bundled with Druid. If false, extensions must be compatible with classes provided by any jars bundled with Druid. If true, extensions may depend on conflicting versions.false
druid.extensions.hadoopContainerDruidClasspathHadoop Indexing launches Hadoop jobs and this configuration provides way to explicitly set the user classpath for the Hadoop job. By default, this is computed automatically by Druid based on the Druid service classpath and set of extensions. However, sometimes you might want to be explicit to resolve dependency conflicts between Druid and Hadoop.null
druid.extensions.addExtensionsToHadoopContainerOnly applicable if druid.extensions.hadoopContainerDruidClasspath is provided. If set to true, then extensions specified in the loadList are added to Hadoop container classpath. Note that when druid.extensions.hadoopContainerDruidClasspath is not provided then extensions are always added to Hadoop container classpath.false

Modules

PropertyDescriptionDefault
druid.modules.excludeListA JSON array of canonical class names (e.g., "org.apache.druid.somepackage.SomeModule") of module classes which shouldn't be loaded, even if they are found in extensions specified by druid.extensions.loadList, or in the list of core modules specified to be loaded on a particular Druid service type. Useful when some useful extension contains some module, which shouldn't be loaded on some Druid service type because some dependencies of that module couldn't be satisfied.[]

ZooKeeper

We recommend just setting the base ZK path and the ZK service host, but all ZK paths that Druid uses can be overwritten to absolute paths.

PropertyDescriptionDefault
druid.zk.paths.baseBase ZooKeeper path./druid
druid.zk.service.hostThe ZooKeeper hosts to connect to. This is a REQUIRED property and therefore a host address must be supplied.none
druid.zk.service.userThe username to authenticate with ZooKeeper. This is an optional property.none
druid.zk.service.pwdThe Password Provider or the string password to authenticate with ZooKeeper. This is an optional property.none
druid.zk.service.authSchemedigest is the only authentication scheme supported.digest

ZooKeeper behavior

PropertyDescriptionDefault
druid.zk.service.sessionTimeoutMsZooKeeper session timeout, in milliseconds.30000
druid.zk.service.connectionTimeoutMsZooKeeper connection timeout, in milliseconds.15000
druid.zk.service.compressBoolean flag for whether or not created Znodes should be compressed.true
druid.zk.service.aclBoolean flag for whether or not to enable ACL security for ZooKeeper. If ACL is enabled, zNode creators will have all permissions.false

Path configuration

Druid interacts with ZooKeeper through a set of standard path configurations. We recommend just setting the base ZooKeeper path, but all ZooKeeper paths that Druid uses can be overwritten to absolute paths.

PropertyDescriptionDefault
druid.zk.paths.baseBase ZooKeeper path./druid
druid.zk.paths.propertiesPathZooKeeper properties path.${druid.zk.paths.base}/properties
druid.zk.paths.announcementsPathDruid service announcement path.${druid.zk.paths.base}/announcements
druid.zk.paths.liveSegmentsPathCurrent path for where Druid services announce their segments.${druid.zk.paths.base}/segments
druid.zk.paths.coordinatorPathUsed by the Coordinator for leader election.${druid.zk.paths.base}/coordinator

The indexing service also uses its own set of paths. These configs can be included in the common configuration.

PropertyDescriptionDefault
druid.zk.paths.indexer.baseBase ZooKeeper path for${druid.zk.paths.base}/indexer
druid.zk.paths.indexer.announcementsPathMiddle Managers announce themselves here.${druid.zk.paths.indexer.base}/announcements
druid.zk.paths.indexer.tasksPathUsed to assign tasks to Middle Managers.${druid.zk.paths.indexer.base}/tasks
druid.zk.paths.indexer.statusPathParent path for announcement of task statuses.${druid.zk.paths.indexer.base}/status

If druid.zk.paths.base and druid.zk.paths.indexer.base are both set, and none of the other druid.zk.paths.* or druid.zk.paths.indexer.* values are set, then the other properties will be evaluated relative to their respective base. For example, if druid.zk.paths.base is set to /druid1 and druid.zk.paths.indexer.base is set to /druid2 then druid.zk.paths.announcementsPath will default to /druid1/announcements while druid.zk.paths.indexer.announcementsPath will default to /druid2/announcements.

The following path is used for service discovery. It is not affected by druid.zk.paths.base and must be specified separately.

PropertyDescriptionDefault
druid.discovery.curator.pathServices announce themselves under this ZooKeeper path./druid/discovery

TLS

General configuration

PropertyDescriptionDefault
druid.enablePlaintextPortEnable/Disable HTTP connector.true
druid.enableTlsPortEnable/Disable HTTPS connector.false

Although not recommended but both HTTP and HTTPS connectors can be enabled at a time and respective ports are configurable using druid.plaintextPort and druid.tlsPort properties on each service. Please see Configuration section of individual services to check the valid and default values for these ports.

Jetty server TLS configuration

Druid uses Jetty as an embedded web server. To learn more about TLS/SSL, certificates, and related concepts in Jetty, including explanations of the configuration settings below, see "Configuring SSL/TLS KeyStores" in the Jetty Operations Guide.

For information about TLS/SSL support in Java in general, see the Java Secure Socket Extension (JSSE) Reference Guide. The Java Cryptography Architecture Standard Algorithm Name Documentation for JDK 8 lists all possible values for the following properties, among others provided by the Java implementation.

PropertyDescriptionDefaultRequired
druid.server.https.keyStorePathThe file path or URL of the TLS/SSL KeyStore.noneyes
druid.server.https.keyStoreTypeThe type of the KeyStore.noneyes
druid.server.https.certAliasAlias of TLS/SSL certificate for the connector.noneyes
druid.server.https.keyStorePasswordThe Password Provider or String password for the KeyStore.noneyes

Following table contains non-mandatory advanced configuration options, use caution.

PropertyDescriptionDefaultRequired
druid.server.https.keyManagerFactoryAlgorithmAlgorithm to use for creating KeyManager, more details here.javax.net.ssl.KeyManagerFactory.getDefaultAlgorithm()no
druid.server.https.keyManagerPasswordThe Password Provider or String password for the Key Manager.noneno
druid.server.https.includeCipherSuitesList of cipher suite names to include. You can either use the exact cipher suite name or a regular expression.Jetty's default include cipher listno
druid.server.https.excludeCipherSuitesList of cipher suite names to exclude. You can either use the exact cipher suite name or a regular expression.Jetty's default exclude cipher listno
druid.server.https.includeProtocolsList of exact protocols names to include.Jetty's default include protocol listno
druid.server.https.excludeProtocolsList of exact protocols names to exclude.Jetty's default exclude protocol listno

Internal client TLS configuration (requires simple-client-sslcontext extension)

These properties apply to the SSLContext that will be provided to the internal HTTP client that Druid services use to communicate with each other. These properties require the simple-client-sslcontext extension to be loaded. Without it, Druid services will be unable to communicate with each other when TLS is enabled.

PropertyDescriptionDefaultRequired
druid.client.https.protocolSSL protocol to use.TLSv1.2no
druid.client.https.trustStoreTypeThe type of the key store where trusted root certificates are stored.java.security.KeyStore.getDefaultType()no
druid.client.https.trustStorePathThe file path or URL of the TLS/SSL Key store where trusted root certificates are stored.noneyes
druid.client.https.trustStoreAlgorithmAlgorithm to be used by TrustManager to validate certificate chainsjavax.net.ssl.TrustManagerFactory.getDefaultAlgorithm()no
druid.client.https.trustStorePasswordThe Password Provider or String password for the Trust Store.noneyes

This document lists all the possible values for the above mentioned configs among others provided by Java implementation.

Authentication and authorization

PropertyTypeDescriptionDefaultRequired
druid.auth.authenticatorChainJSON List of StringsList of Authenticator type names["allowAll"]no
druid.escalator.typeStringType of the Escalator that should be used for internal Druid communications. This Escalator must use an authentication scheme that is supported by an Authenticator in druid.auth.authenticatorChain.noopno
druid.auth.authorizersJSON List of StringsList of Authorizer type names["allowAll"]no
druid.auth.unsecuredPathsList of StringsList of paths for which security checks will not be performed. All requests to these paths will be allowed.[]no
druid.auth.allowUnauthenticatedHttpOptionsBooleanIf true, skip authentication checks for HTTP OPTIONS requests. This is needed for certain use cases, such as supporting CORS pre-flight requests. Note that disabling authentication checks for OPTIONS requests will allow unauthenticated users to determine what Druid endpoints are valid (by checking if the OPTIONS request returns a 200 instead of 404), so enabling this option may reveal information about server configuration, including information about what extensions are loaded (if those extensions add endpoints).falseno

For more information, please see Authentication and Authorization.

For configuration options for specific auth extensions, please refer to the extension documentation.

Startup logging

All services can log debugging information on startup.

PropertyDescriptionDefault
druid.startup.logging.logPropertiesLog all properties on startup (from common.runtime.properties, runtime.properties, and the JVM command line).false
druid.startup.logging.maskPropertiesMasks sensitive properties (passwords, for example) containing theses words.["password"]

Note that some sensitive information may be logged if these settings are enabled.

Request logging

All services that can serve queries can also log the query requests they see. Broker services can additionally log the SQL requests (both from HTTP and JDBC) they see. For an example of setting up request logging, see Request logging.

PropertyDescriptionDefault
druid.request.logging.typeHow to log every query request. Choices: noop, file, emitter, slf4j, filtered, composing, switchingnoop (request logging disabled by default)

To enable sending all the HTTP requests to a log, set org.apache.druid.jetty.RequestLog to the DEBUG level. See Logging for more information.

File request logging

The file request logger stores daily request logs on disk.

PropertyDescriptionDefault
druid.request.logging.dirHistorical, Realtime, and Broker services maintain request logs of all of the requests they get (interaction is via POST, so normal request logs don’t generally capture information about the actual query), this specifies the directory to store the request logs innone
druid.request.logging.filePatternJoda datetime format for each file"yyyy-MM-dd'.log'"
druid.request.logging.durationToRetainPeriod to retain the request logs on disk. The period should be at least longer than P1D.none

The format of request logs is TSV, one line per requests, with five fields: timestamp, remote_addr, native_query, query_context, sql_query.

For native JSON request, the sql_query field is empty. For example:

2019-01-14T10:00:00.000Z        127.0.0.1   {"queryType":"topN","dataSource":{"type":"table","name":"wikiticker"},"virtualColumns":[],"dimension":{"type":"LegacyDimensionSpec","dimension":"page","outputName":"page","outputType":"STRING"},"metric":{"type":"LegacyTopNMetricSpec","metric":"count"},"threshold":10,"intervals":{"type":"LegacySegmentSpec","intervals":["2015-09-12T00:00:00.000Z/2015-09-13T00:00:00.000Z"]},"filter":null,"granularity":{"type":"all"},"aggregations":[{"type":"count","name":"count"}],"postAggregations":[],"context":{"queryId":"74c2d540-d700-4ebd-b4a9-3d02397976aa"},"descending":false}    {"query/time":100,"query/bytes":800,"success":true,"identity":"user1"}

For SQL query request, the native_query field is empty. For example:

2019-01-14T10:00:00.000Z        127.0.0.1       {"sqlQuery/time":100, "sqlQuery/planningTimeMs":10, "sqlQuery/bytes":600, "success":true, "identity":"user1"}  {"query":"SELECT page, COUNT(*) AS Edits FROM wikiticker WHERE TIME_IN_INTERVAL(\"__time\", '2015-09-12/2015-09-13') GROUP BY page ORDER BY Edits DESC LIMIT 10","context":{"sqlQueryId":"c9d035a0-5ffd-4a79-a865-3ffdadbb5fdd","nativeQueryIds":"[490978e4-f5c7-4cf6-b174-346e63cf8863]"}}

Emitter request logging

The emitter request logger emits every request to the external location specified in the emitter configuration.

PropertyDescriptionDefault
druid.request.logging.feedFeed name for requests.none

SLF4J request logging

The slf4j request logger logs every request using SLF4J. It serializes native queries into JSON in the log message regardless of the SLF4J format specification. Requests are logged under the class org.apache.druid.server.log.LoggingRequestLogger.

PropertyDescriptionDefault
druid.request.logging.setMDCIf you want to set MDC entries within the log entry, set this value to true. Your logging system must be configured to support MDC in order to format this data.false
druid.request.logging.setContextMDCSet to "true" to add the Druid query context to the MDC entries. Only applies when setMDC is true.false

For a native query, the following MDC fields are populated when setMDC is true:

MDC fieldDescription
queryIdThe query ID
sqlQueryIdThe SQL query ID if this query is part of a SQL request
dataSourceThe datasource the query was against
queryTypeThe type of the query
hasFiltersIf the query has any filters
remoteAddrThe remote address of the requesting client
durationThe duration of the query interval
resultOrderingThe ordering of results
descendingIf the query is a descending query

Filtered request logging

The filtered request logger filters requests based on the query type or how long a query takes to complete. For native queries, the logger only logs requests when the query/time metric exceeds the threshold provided in queryTimeThresholdMs. For SQL queries, it only logs requests when the sqlQuery/time metric exceeds threshold provided in sqlQueryTimeThresholdMs. See Metrics for more details on query metrics.

Requests that meet the threshold are logged using the request logger type set in druid.request.logging.delegate.type.

PropertyDescriptionDefault
druid.request.logging.queryTimeThresholdMsThreshold value for the query/time metric in milliseconds.0, i.e., no filtering
druid.request.logging.sqlQueryTimeThresholdMsThreshold value for the sqlQuery/time metric in milliseconds.0, i.e., no filtering
druid.request.logging.mutedQueryTypesQuery requests of these types are not logged. Query types are defined as string objects corresponding to the "queryType" value for the specified query in the Druid's native JSON query API. Misspelled query types will be ignored. Example to ignore scan and timeBoundary queries: ["scan", "timeBoundary"][]
druid.request.logging.delegate.typeType of delegate request logger to log requests.none

Composing request logging

The composing request logger emits request logs to multiple request loggers.

PropertyDescriptionDefault
druid.request.logging.loggerProvidersList of request loggers for emitting request logs.none

Switching request logging

The switching request logger routes native query request logs to one request logger and SQL query request logs to another request logger.

PropertyDescriptionDefault
druid.request.logging.nativeQueryLoggerRequest logger for emitting native query request logs.none
druid.request.logging.sqlQueryLoggerRequest logger for emitting SQL query request logs.none

Audit logging

Coordinator and Overlord log changes to lookups, segment load/drop rules, and dynamic configuration changes for auditing.

PropertyDescriptionDefault
druid.audit.manager.typeType of audit manager used for handling audited events. Audited events are logged when set to log or persisted in metadata store when set to sql.sql
druid.audit.manager.logLevelLog level of audit events with possible values DEBUG, INFO, WARN. This property is used only when druid.audit.manager.type is set to log.INFO
druid.audit.manager.auditHistoryMillisDefault duration for querying audit history.1 week
druid.audit.manager.includePayloadAsDimensionInMetricBoolean flag on whether to add payload column in service metric.false
druid.audit.manager.maxPayloadSizeBytesThe maximum size of audit payload to store in Druid's metadata store audit table. If the size of audit payload exceeds this value, the audit log would be stored with a message indicating that the payload was omitted instead. Setting maxPayloadSizeBytes to -1 (default value) disables this check, meaning Druid will always store audit payload regardless of it's size. Setting to any negative number other than -1 is invalid. Human-readable format is supported, see here.-1
druid.audit.manager.skipNullFieldIf true, the audit payload stored in metadata store will exclude any field with null value.false

Enabling metrics

You can configure Druid services to emit metrics regularly from a number of monitors via emitters.

PropertyDescriptionDefault
druid.monitoring.emissionPeriodFrequency that Druid emits metrics.PT1M
druid.monitoring.monitorsSets list of Druid monitors used by a service.

Because individual monitors sometimes only work on specific process types, it is best to set this property for each process type individually in e.g. historical/runtime.properties rather than _common/common.runtime.properties.
none (no monitors)
druid.emitterSetting this value initializes one of the emitter modules.noop (metric emission disabled by default)

Metrics monitors

Metric monitoring is an essential part of Druid operations. The following monitors are available:

NameDescription
org.apache.druid.client.cache.CacheMonitorEmits metrics (to logs) about the segment results cache for Historical and Broker services. Reports typical cache statistics include hits, misses, rates, and size (bytes and number of entries), as well as timeouts and and errors.
org.apache.druid.java.util.metrics.SysMonitorReports on various system activities and statuses using the SIGAR library. Requires execute privileges on files in java.io.tmpdir. Do not set java.io.tmpdir to noexec when using SysMonitor.
org.apache.druid.java.util.metrics.JvmMonitorReports various JVM-related statistics.
org.apache.druid.java.util.metrics.JvmCpuMonitorReports statistics of CPU consumption by the JVM.
org.apache.druid.java.util.metrics.CpuAcctDeltaMonitorReports consumed CPU as per the cpuacct cgroup.
org.apache.druid.java.util.metrics.JvmThreadsMonitorReports Thread statistics in the JVM, like numbers of total, daemon, started, died threads.
org.apache.druid.java.util.metrics.CgroupCpuMonitorReports CPU shares and quotas as per the cpu cgroup.
org.apache.druid.java.util.metrics.CgroupCpuSetMonitorReports CPU core/HT and memory node allocations as per the cpuset cgroup.
org.apache.druid.java.util.metrics.CgroupDiskMonitorReports disk statistic as per the blkio cgroup.
org.apache.druid.java.util.metrics.CgroupMemoryMonitorReports memory statistic as per the memory cgroup.
org.apache.druid.server.metrics.HistoricalMetricsMonitorReports statistics on Historical services. Available only on Historical services.
org.apache.druid.server.metrics.SegmentStatsMonitorEXPERIMENTAL Reports statistics about segments on Historical services. Available only on Historical services. Not to be used when lazy loading is configured.
org.apache.druid.server.metrics.QueryCountStatsMonitorReports how many queries have been successful/failed/interrupted.
org.apache.druid.server.metrics.SubqueryCountStatsMonitorReports how many subqueries have been materialized as rows or bytes and various other statistics related to the subquery execution
org.apache.druid.server.emitter.HttpEmittingMonitorReports internal metrics of http or parametrized emitter (see below). Must not be used with another emitter type. See the description of the metrics here: https://github.com/apache/druid/pull/4973.
org.apache.druid.server.metrics.TaskCountStatsMonitorReports how many ingestion tasks are currently running/pending/waiting and also the number of successful/failed tasks per emission period.
org.apache.druid.server.metrics.TaskSlotCountStatsMonitorReports metrics about task slot usage per emission period.
org.apache.druid.server.metrics.WorkerTaskCountStatsMonitorReports how many ingestion tasks are currently running/pending/waiting, the number of successful/failed tasks, and metrics about task slot usage for the reporting worker, per emission period. Only supported by Middle Manager node types.
org.apache.druid.server.metrics.ServiceStatusMonitorReports a heartbeat for the service.

For example, you might configure monitors on all services for system and JVM information within common.runtime.properties as follows:

druid.monitoring.monitors=["org.apache.druid.java.util.metrics.SysMonitor","org.apache.druid.java.util.metrics.JvmMonitor"]

You can override cluster-wide configuration by amending the runtime.properties of individual services.

Metrics emitters

There are several emitters available:

  • noop (default) disables metric emission.
  • logging emits logs using Log4j2.
  • http sends POST requests of JSON events.
  • parametrized operates like the http emitter but fine-tunes the recipient URL based on the event feed.
  • composing initializes multiple emitter modules.
  • graphite emits metrics to a Graphite Carbon service.
  • switching initializes and emits to multiple emitter modules based on the event feed.
Logging emitter module

The use this emitter module, set druid.emitter=logging. The logging emitter uses a Log4j2 logger named druid.emitter.logging.loggerClass to emit events. Each event is logged as a single json object with a Marker as the feed of the event. Users may wish to edit the log4j config to route these logs to different sources based on the feed of the event.

PropertyDescriptionDefault
druid.emitter.logging.loggerClassThe class used for logging.org.apache.druid.java.util.emitter.core.LoggingEmitter
druid.emitter.logging.logLevelChoices: debug, info, warn, error. The log level at which message are logged.info
HTTP emitter module
PropertyDescriptionDefault
druid.emitter.http.flushMillisHow often the internal message buffer is flushed (data is sent).60000
druid.emitter.http.flushCountHow many messages the internal message buffer can hold before flushing (sending).500
druid.emitter.http.basicAuthenticationPassword Provider for providing login and password for authentication in "login:password" form. For example, druid.emitter.http.basicAuthentication=admin:adminpassword uses Default Password Provider which allows plain text passwords.not specified = no authentication
druid.emitter.http.flushTimeOutThe timeout after which an event should be sent to the endpoint, even if internal buffers are not filled, in milliseconds.not specified = no timeout
druid.emitter.http.batchingStrategyThe strategy of how the batch is formatted. "ARRAY" means [event1,event2], "NEWLINES" means event1\nevent2, ONLY_EVENTS means event1event2.ARRAY
druid.emitter.http.maxBatchSizeThe maximum batch size, in bytes.the minimum of (10% of JVM heap size divided by 2) or (5242880 (i. e. 5 MiB))
druid.emitter.http.batchQueueSizeLimitThe maximum number of batches in emitter queue, if there are problems with emitting.the maximum of (2) or (10% of the JVM heap size divided by 5MiB)
druid.emitter.http.minHttpTimeoutMillisIf the speed of filling batches imposes timeout smaller than that, not even trying to send batch to endpoint, because it will likely fail, not being able to send the data that fast. Configure this depending based on emitter/successfulSending/minTimeMs metric. Reasonable values are 10ms..100ms.0
druid.emitter.http.recipientBaseUrlThe base URL to emit messages to. Druid will POST JSON to be consumed at the HTTP endpoint specified by this property.none, required config
HTTP emitter module TLS overrides

By default, when sending events to a TLS-enabled receiver, the HTTP Emitter uses an SSLContext obtained from the service described at Druid's internal communication over TLS, that is the same SSLContext that would be used for internal communications between Druid services.

In some use cases it may be desirable to have the HTTP Emitter use its own separate truststore configuration. For example, there may be organizational policies that prevent the TLS-enabled metrics receiver's certificate from being added to the same truststore used by Druid's internal HTTP client.

The following properties allow the HTTP Emitter to use its own truststore configuration when building its SSLContext.

PropertyDescriptionDefault
druid.emitter.http.ssl.useDefaultJavaContextIf set to true, the HttpEmitter will use SSLContext.getDefault(), the default Java SSLContext, and all other properties below are ignored.false
druid.emitter.http.ssl.trustStorePathThe file path or URL of the TLS/SSL Key store where trusted root certificates are stored. If this is unspecified, the HTTP Emitter will use the same SSLContext as Druid's internal HTTP client, as described in the beginning of this section, and all other properties below are ignored.null
druid.emitter.http.ssl.trustStoreTypeThe type of the key store where trusted root certificates are stored.java.security.KeyStore.getDefaultType()
druid.emitter.http.ssl.trustStoreAlgorithmAlgorithm to be used by TrustManager to validate certificate chainsjavax.net.ssl.TrustManagerFactory.getDefaultAlgorithm()
druid.emitter.http.ssl.trustStorePasswordThe Password Provider or String password for the Trust Store.none
druid.emitter.http.ssl.protocolTLS protocol to use."TLSv1.2"
Parametrized HTTP emitter module

The parametrized emitter takes the same configs as the http emitter using the prefix druid.emitter.parametrized.httpEmitting.. For example:

  • druid.emitter.parametrized.httpEmitting.flushMillis
  • druid.emitter.parametrized.httpEmitting.flushCount
  • druid.emitter.parametrized.httpEmitting.ssl.trustStorePath

Do not specify recipientBaseUrl with the parametrized emitter. Instead use recipientBaseUrlPattern described in the table below.

PropertyDescriptionDefault
druid.emitter.parametrized.recipientBaseUrlPatternThe URL pattern to send an event to, based on the event's feed. For example, http://foo.bar/{feed}, that will send event to http://foo.bar/metrics if the event's feed is "metrics".none, required config
Composing emitter module
PropertyDescriptionDefault
druid.emitter.composing.emittersList of emitter modules to load, such as ["logging","http"].[]
Graphite emitter

To use graphite as emitter set druid.emitter=graphite. For configuration details, see Graphite emitter for the Graphite emitter Druid extension.

Switching emitter

To use switching as emitter set druid.emitter=switching.

PropertyDescriptionDefault
druid.emitter.switching.emittersJSON map of feed to list of emitter modules that will be used for the mapped feed, such as {"metrics":["http"], "alerts":["logging"]}{}
druid.emitter.switching.defaultEmittersJSON list of emitter modules to load that will be used if there is no emitter specifically designated for that event's feed, such as ["logging","http"].[]

Metadata storage

These properties specify the JDBC connection and other configuration around the metadata storage. The only services that connect to the metadata storage with these properties are the Coordinator and Overlord.

PropertyDescriptionDefault
druid.metadata.storage.typeThe type of metadata storage to use. One of mysql, postgresql, or derby.derby
druid.metadata.storage.connector.connectURIThe JDBC URI for the database to connect tonone
druid.metadata.storage.connector.userThe username to connect with.none
druid.metadata.storage.connector.passwordThe Password Provider or String password used to connect with.none
druid.metadata.storage.connector.createTablesIf Druid requires a table and it doesn't exist, create it?true
druid.metadata.storage.tables.baseThe base name for tables.druid
druid.metadata.storage.tables.dataSourceThe table to use to look for datasources created by Kafka Indexing Service.druid_dataSource
druid.metadata.storage.tables.pendingSegmentsThe table to use to look for pending segments.druid_pendingSegments
druid.metadata.storage.tables.segmentsThe table to use to look for segments.druid_segments
druid.metadata.storage.tables.rulesThe table to use to look for segment load/drop rules.druid_rules
druid.metadata.storage.tables.configThe table to use to look for configs.druid_config
druid.metadata.storage.tables.tasksUsed by the indexing service to store tasks.druid_tasks
druid.metadata.storage.tables.taskLogUsed by the indexing service to store task logs.druid_tasklogs
druid.metadata.storage.tables.taskLockUsed by the indexing service to store task locks.druid_tasklocks
druid.metadata.storage.tables.supervisorsUsed by the indexing service to store supervisor configurations.druid_supervisors
druid.metadata.storage.tables.auditThe table to use for audit history of configuration changes, such as Coordinator rules.druid_audit

Deep storage

The configurations concern how to push and pull Segments from deep storage.

PropertyDescriptionDefault
druid.storage.typeThe type of deep storage to use. One of local, noop, s3, hdfs, c*.local

Local deep storage

Local deep storage uses the local filesystem.

PropertyDescriptionDefault
druid.storage.storageDirectoryDirectory on disk to use as deep storage./tmp/druid/localStorage

Noop deep storage

This deep storage doesn't do anything. There are no configs.

S3 deep storage

This deep storage is used to interface with Amazon's S3. Note that the druid-s3-extensions extension must be loaded. The below table shows some important configurations for S3. See S3 Deep Storage for full configurations.

PropertyDescriptionDefault
druid.storage.bucketS3 bucket name.none
druid.storage.baseKeyS3 object key prefix for storage.none
druid.storage.disableAclBoolean flag for ACL. If this is set to false, the full control would be granted to the bucket owner. This may require to set additional permissions. See S3 permissions settings.false
druid.storage.archiveBucketS3 bucket name for archiving when running the archive task.none
druid.storage.archiveBaseKeyS3 object key prefix for archiving.none
druid.storage.sse.typeServer-side encryption type. Should be one of s3, kms, and custom. See the below Server-side encryption section for more details.None
druid.storage.sse.kms.keyIdAWS KMS key ID. This is used only when druid.storage.sse.type is kms and can be empty to use the default key ID.None
druid.storage.sse.custom.base64EncodedKeyBase64-encoded key. Should be specified if druid.storage.sse.type is custom.None
druid.storage.useS3aSchemaIf true, use the "s3a" filesystem when using Hadoop-based ingestion. If false, the "s3n" filesystem will be used. Only affects Hadoop-based ingestion.false

HDFS deep storage

This deep storage is used to interface with HDFS. You must load the druid-hdfs-storage extension.

PropertyDescriptionDefault
druid.storage.storageDirectoryHDFS directory to use as deep storage.none

Cassandra deep storage

This deep storage is used to interface with Cassandra. You must load the druid-cassandra-storage extension.

PropertyDescriptionDefault
druid.storage.hostCassandra host.none
druid.storage.keyspaceCassandra key space.none

Centralized datasource schema

Centralized datasource schema is an experimental feature to centralized datasource schema building within the Coordinator. Traditionally, the datasource schema is built in the Brokers by combining schema for all the available segments of a datasource. Brokers issue segment metadata query to data nodes and tasks to fetch segment schema. In the new arrangement, tasks publish segment schema along with segment metadata to the database and schema for realtime segments is periodically pushed to the Coordinator in the segment announcement flow. This enables Coordinator to cache segment schemas and build datasource schema by combining segment schema. Brokers query the datasource schema from the Coordinator, while retaining the ability to build table schema if the need arises.

PropertyDescriptionDefaultRequired
druid.centralizedDatasourceSchema.enabledBoolean flag for enabling datasource schema building in the Coordinator, this should be specified in the common runtime properties.falseNo.
druid.indexer.fork.property.druid.centralizedDatasourceSchema.enabledThis config should be set when CentralizedDatasourceSchema feature is enabled. This should be specified in the Middle Manager runtime properties.falseNo.

If you enable this feature, you can query datasources that are only stored in deep storage and are not loaded on a Historical. For more information, see Query from deep storage.

For stale schema cleanup configs, refer to properties with the prefix druid.coordinator.kill.segmentSchema in Metadata Management.

Ingestion security configuration

HDFS input source

You can set the following property to specify permissible protocols for the HDFS input source.

PropertyPossible valuesDescriptionDefault
druid.ingestion.hdfs.allowedProtocolsList of protocolsAllowed protocols for the HDFS input source.["hdfs"]

HTTP input source

You can set the following property to specify permissible protocols for the HTTP input source.

PropertyPossible valuesDescriptionDefault
druid.ingestion.http.allowedProtocolsList of protocolsAllowed protocols for the HTTP input source.["http", "https"]

External data access security configuration

JDBC connections to external databases

You can use the following properties to specify permissible JDBC options for:

These properties do not apply to metadata storage connections.

PropertyPossible valuesDescriptionDefault
druid.access.jdbc.enforceAllowedPropertiesBooleanWhen true, Druid applies druid.access.jdbc.allowedProperties to JDBC connections starting with jdbc:postgresql:, jdbc:mysql:, or jdbc:mariadb:. When false, Druid allows any kind of JDBC connections without JDBC property validation. This config is for backward compatibility especially during upgrades since enforcing allow list can break existing ingestion jobs or lookups based on JDBC. This config is deprecated and will be removed in a future release.true
druid.access.jdbc.allowedPropertiesList of JDBC propertiesDefines a list of allowed JDBC properties. Druid always enforces the list for all JDBC connections starting with jdbc:postgresql:, jdbc:mysql:, and jdbc:mariadb: if druid.access.jdbc.enforceAllowedProperties is set to true.

This option is tested against MySQL connector 8.2.0, MariaDB connector 2.7.4, and PostgreSQL connector 42.2.14. Other connector versions might not work.
["useSSL", "requireSSL", "ssl", "sslmode"]
druid.access.jdbc.allowUnknownJdbcUrlFormatBooleanWhen false, Druid only accepts JDBC connections starting with jdbc:postgresql: or jdbc:mysql:. When true, Druid allows JDBC connections to any kind of database, but only enforces druid.access.jdbc.allowedProperties for PostgreSQL and MySQL/MariaDB.true

Task logging

You can use the druid.indexer configuration to set a long-term storage location for task log files, and to set a retention policy.

For more information about ingestion tasks and the services of generating logs, see the task reference.

Log long-term storage

PropertyDescriptionDefault
druid.indexer.logs.typeWhere to store task logs. noop, s3, azure, google, hdfs, filefile
File task logs

Store task logs in the local filesystem.

PropertyDescriptionDefault
druid.indexer.logs.directoryLocal filesystem path.log
S3 task logs

Store task logs in S3. Note that the druid-s3-extensions extension must be loaded.

PropertyDescriptionDefault
druid.indexer.logs.s3BucketS3 bucket name.none
druid.indexer.logs.s3PrefixS3 key prefix.none
druid.indexer.logs.disableAclBoolean flag for ACL. If this is set to false, the full control would be granted to the bucket owner. If the task logs bucket is the same as the deep storage (S3) bucket, then the value of this property will need to be set to true if druid.storage.disableAcl has been set to true.false
Azure Blob Store task logs

Store task logs in Azure Blob Store. To enable this feature, load the druid-azure-extensions extension, and configure deep storage for Azure. Druid uses the same authentication method configured for deep storage and stores task logs in the same storage account (set in druid.azure.account).

PropertyDescriptionDefault
druid.indexer.logs.containerThe Azure Blob Store container to write logs to.Must be set.
druid.indexer.logs.prefixThe path to prepend to logs.Must be set.
Google Cloud Storage task logs

Store task logs in Google Cloud Storage.

Note: The druid-google-extensions extension must be loaded, and this uses the same storage settings as the deep storage module for google.

PropertyDescriptionDefault
druid.indexer.logs.bucketThe Google Cloud Storage bucket to write logs tonone
druid.indexer.logs.prefixThe path to prepend to logsnone
HDFS task logs

Store task logs in HDFS. Note that the druid-hdfs-storage extension must be loaded.

PropertyDescriptionDefault
druid.indexer.logs.directoryThe directory to store logs.none

Log retention policy

PropertyDescriptionDefault
druid.indexer.logs.kill.enabledBoolean value for whether to enable deletion of old task logs. If set to true, Overlord will submit kill tasks periodically based on druid.indexer.logs.kill.delay specified, which will delete task logs from the log directory as well as tasks and tasklogs table entries in metadata storage except for tasks created in the last druid.indexer.logs.kill.durationToRetain period.false
druid.indexer.logs.kill.durationToRetainRequired if kill is enabled. In milliseconds, task logs and entries in task-related metadata storage tables to be retained created in last x milliseconds.None
druid.indexer.logs.kill.initialDelayOptional. Number of milliseconds after Overlord start when first auto kill is run.random value less than 300000 (5 mins)
druid.indexer.logs.kill.delayOptional. Number of milliseconds of delay between successive executions of auto kill run.21600000 (6 hours)

API error response

You can configure Druid API error responses to hide internal information like the Druid class name, stack trace, thread name, servlet name, code, line/column number, host, or IP address.

PropertyDescriptionDefault
druid.server.http.showDetailedJettyErrorsWhen set to true, any error from the Jetty layer / Jetty filter includes the following fields in the JSON response: servlet, message, url, status, and cause, if it exists. When set to false, the JSON response only includes message, url, and status. The field values remain unchanged.true
druid.server.http.errorResponseTransform.strategyError response transform strategy. The strategy controls how Druid transforms error responses from Druid services. When unset or set to none, Druid leaves error responses unchanged.none

Error response transform strategy

You can use an error response transform strategy to transform error responses from within Druid services to hide internal information. When you specify an error response transform strategy other than none, Druid transforms the error responses from Druid services as follows:

  • For any query API that fails in the Router service, Druid sets the fields errorClass and host to null. Druid applies the transformation strategy to the errorMessage field.
  • For any SQL query API that fails, for example POST /druid/v2/sql/..., Druid sets the fields errorClass and host to null. Druid applies the transformation strategy to the errorMessage field.
  • For any JDBC related exceptions, Druid will turn all checked exceptions into QueryInterruptedException otherwise druid will attempt to keep the exception as the same type. For example if the original exception isn't owned by Druid it will become QueryInterruptedException. Druid applies the transformation strategy to the errorMessage field.
No error response transform strategy

In this mode, Druid leaves error responses from underlying services unchanged and returns the unchanged errors to the API client. This is the default Druid error response mode. To explicitly enable this strategy, set druid.server.http.errorResponseTransform.strategy to none.

Allowed regular expression error response transform strategy

In this mode, Druid validates the error responses from underlying services against a list of regular expressions. Only error messages that match a configured regular expression are returned. To enable this strategy, set druid.server.http.errorResponseTransform.strategy to allowedRegex.

PropertyDescriptionDefault
druid.server.http.errorResponseTransform.allowedRegexThe list of regular expressions Druid uses to validate error messages. If the error message matches any of the regular expressions, then Druid includes it in the response unchanged. If the error message does not match any of the regular expressions, Druid replaces the error message with null or with a default message depending on the type of underlying Exception.[]

For example, consider the following error response:

{"error":"Plan validation failed","errorMessage":"org.apache.calcite.runtime.CalciteContextException: From line 1, column 15 to line 1, column 38: Object 'nonexistent-datasource' not found","errorClass":"org.apache.calcite.tools.ValidationException","host":null}

If druid.server.http.errorResponseTransform.allowedRegex is set to [], Druid transforms the query error response to the following:

{"error":"Plan validation failed","errorMessage":null,"errorClass":null,"host":null}

On the other hand, if druid.server.http.errorResponseTransform.allowedRegex is set to [".*CalciteContextException.*"] then Druid transforms the query error response to the following:

{"error":"Plan validation failed","errorMessage":"org.apache.calcite.runtime.CalciteContextException: From line 1, column 15 to line 1, column 38: Object 'nonexistent-datasource' not found","errorClass":null,"host":null}

Overlord discovery

This config is used to find the Overlord using Curator service discovery. Only required if you are actually running an Overlord.

PropertyDescriptionDefault
druid.selectors.indexing.serviceNameThe druid.service name of the Overlord service. To start the Overlord with a different name, set it with this property.druid/overlord

Coordinator discovery

This config is used to find the Coordinator using Curator service discovery. This config is used by the realtime indexing services to get information about the segments loaded in the cluster.

PropertyDescriptionDefault
druid.selectors.coordinator.serviceNameThe druid.service name of the Coordinator service. To start the Coordinator with a different name, set it with this property.druid/coordinator

Announcing segments

You can configure how to announce and unannounce Znodes in ZooKeeper (using Curator). For normal operations you do not need to override any of these configs.

Batch data segment announcer

In current Druid, multiple data segments may be announced under the same Znode.

PropertyDescriptionDefault
druid.announcer.segmentsPerNodeEach Znode contains info for up to this many segments.50
druid.announcer.maxBytesPerNodeMax byte size for Znode. Allowed range is [1024, 1048576].524288
druid.announcer.skipDimensionsAndMetricsSkip Dimensions and Metrics list from segment announcements. NOTE: Enabling this will also remove the dimensions and metrics list from Coordinator and Broker endpoints.false
druid.announcer.skipLoadSpecSkip segment LoadSpec from segment announcements. NOTE: Enabling this will also remove the loadspec from Coordinator and Broker endpoints.false

If you want to turn off the batch data segment announcer, you can add a property to skip announcing segments. You do not want to enable this config if you have any services using batch for druid.serverview.type

PropertyDescriptionDefault
druid.announcer.skipSegmentAnnouncementOnZkSkip announcing segments to ZooKeeper. Note that the batch server view will not work if this is set to true.false

JavaScript

Druid supports dynamic runtime extension through JavaScript functions. This functionality can be configured through the following properties.

PropertyDescriptionDefault
druid.javascript.enabledSet to "true" to enable JavaScript functionality. This affects the JavaScript parser, filter, extractionFn, aggregator, post-aggregator, router strategy, and worker selection strategy.false
info

JavaScript-based functionality is disabled by default. Please refer to the Druid JavaScript programming guide for guidelines about using Druid's JavaScript functionality, including instructions on how to enable it.

Double column storage

Prior to version 0.13.0, Druid's storage layer used a 32-bit float representation to store columns created by the doubleSum, doubleMin, and doubleMax aggregators at indexing time. Starting from version 0.13.0 the default will be 64-bit floats for Double columns. Using 64-bit representation for double column will lead to avoid precision loss at the cost of doubling the storage size of such columns. To keep the old format set the system-wide property druid.indexing.doubleStorage=float. You can also use floatSum, floatMin, and floatMax to use 32-bit float representation. Support for 64-bit floating point columns was released in Druid 0.11.0, so if you use this feature then older versions of Druid will not be able to read your data segments.

PropertyDescriptionDefault
druid.indexing.doubleStorageSet to "float" to use 32-bit double representation for double columns.double

SQL compatible null handling

These configurations are deprecated and will be removed in a future release at which point Druid will always have SQl compatible null handling.

Prior to version 0.13.0, Druid string columns treated '' and null values as interchangeable, and numeric columns were unable to represent null values, coercing null to 0. Druid 0.13.0 introduced a mode which enabled SQL compatible null handling, allowing string columns to distinguish empty strings from nulls, and numeric columns to contain null rows.

PropertyDescriptionDefault
druid.generic.useDefaultValueForNullSet to false to store and query data in SQL compatible mode. This configuration has been deprecated and will be removed in a future release, taking on the false behavior. When set to true (deprecated legacy mode), null values will be stored as '' for string columns and 0 for numeric columns.false
druid.generic.useThreeValueLogicForNativeFiltersSet to true to use SQL compatible three-value logic when processing native Druid filters when druid.generic.useDefaultValueForNull=false and druid.expressions.useStrictBooleans=true. This configuration has been deprecated and will be removed in a future release, taking on the true behavior. When set to false Druid uses 2 value logic for filter processing, even when druid.generic.useDefaultValueForNull=false and druid.expressions.useStrictBooleans=true. See boolean handling for more detailstrue
druid.generic.ignoreNullsForStringCardinalityWhen set to true, null values will be ignored for the built-in cardinality aggregator over string columns. Set to false to include null values while estimating cardinality of only string columns using the built-in cardinality aggregator. This setting takes effect only when druid.generic.useDefaultValueForNull is set to true and is ignored in SQL compatibility mode. Additionally, empty strings (equivalent to null) are not counted when this is set to true. This configuration has been deprecated and will be removed in a future release since it has no effect when druid.generic.useDefaultValueForNull=false.false

This mode does have a storage size and query performance cost, see segment documentation for more details.

HTTP client

All Druid components can communicate with each other over HTTP.

PropertyDescriptionDefault
druid.global.http.numConnectionsSize of connection pool per destination URL. If there are more HTTP requests than this number that all need to speak to the same URL, then they will queue up.20
druid.global.http.eagerInitializationIndicates that http connections should be eagerly initialized. If set to true, numConnections connections are created upon initializationfalse
druid.global.http.compressionCodecCompression codec to communicate with others. May be "gzip" or "identity".gzip
druid.global.http.readTimeoutThe timeout for data reads.PT15M
druid.global.http.unusedConnectionTimeoutThe timeout for idle connections in connection pool. The connection in the pool will be closed after this timeout and a new one will be established. This timeout should be less than druid.global.http.readTimeout. Set this timeout = ~90% of druid.global.http.readTimeoutPT4M
druid.global.http.numMaxThreadsMaximum number of I/O worker threadsmax(10, ((number of cores * 17) / 16 + 2) + 30)
druid.global.http.clientConnectTimeoutThe timeout (in milliseconds) for establishing client connections.500

Common endpoints configuration

This section contains the configuration options for endpoints that are supported by all services.

PropertyDescriptionDefault
druid.server.hiddenPropertiesIf property names or substring of property names (case insensitive) is in this list, responses of the /status/properties endpoint do not show these properties["druid.s3.accessKey","druid.s3.secretKey","druid.metadata.storage.connector.password", "password", "key", "token", "pwd"]

Master server

This section contains the configuration options for the services that reside on Master servers (Coordinators and Overlords) in the suggested three-server configuration.

Coordinator

For general Coordinator services information, see Coordinator service.

Static Configuration

These Coordinator static configurations can be defined in the coordinator/runtime.properties file.

Coordinator service config
PropertyDescriptionDefault
druid.hostThe host for the current service. This is used to advertise the current service location as reachable from another service and should generally be specified such that http://${druid.host}/ could actually talk to this service.InetAddress.getLocalHost().getCanonicalHostName()
druid.bindOnHostIndicating whether the service's internal jetty server bind on druid.host. Default is false, which means binding to all interfaces.false
druid.plaintextPortThis is the port to actually listen on; unless port mapping is used, this will be the same port as is on druid.host8081
druid.tlsPortTLS port for HTTPS connector, if druid.enableTlsPort is set then this config will be used. If druid.host contains port then that port will be ignored. This should be a non-negative integer.8281
druid.serviceThe name of the service. This is used as a dimension when emitting metrics and alerts to differentiate between the various services.druid/coordinator
Coordinator operation
PropertyDescriptionDefault
druid.coordinator.periodThe run period for the Coordinator. The Coordinator operates by maintaining the current state of the world in memory and periodically looking at the set of "used" segments and segments being served to make decisions about whether any changes need to be made to the data topology. This property sets the delay between each of these runs.PT60S
druid.coordinator.period.indexingPeriodHow often to send compact/merge/conversion tasks to the indexing service. It's recommended to be longer than druid.manager.segments.pollDurationPT1800S (30 mins)
druid.coordinator.startDelayThe operation of the Coordinator works on the assumption that it has an up-to-date view of the state of the world when it runs, the current ZooKeeper interaction code, however, is written in a way that doesn’t allow the Coordinator to know for a fact that it’s done loading the current state of the world. This delay is a hack to give it enough time to believe that it has all the data.PT300S
druid.coordinator.load.timeoutThe timeout duration for when the Coordinator assigns a segment to a Historical service.PT15M
druid.coordinator.kill.pendingSegments.onBoolean flag for whether or not the Coordinator clean up old entries in the pendingSegments table of metadata store. If set to true, Coordinator will check the created time of most recently complete task. If it doesn't exist, it finds the created time of the earliest running/pending/waiting tasks. Once the created time is found, then for all datasources not in the killPendingSegmentsSkipList (see Dynamic configuration), Coordinator will ask the Overlord to clean up the entries 1 day or more older than the found created time in the pendingSegments table. This will be done periodically based on druid.coordinator.period.indexingPeriod specified.true
druid.coordinator.kill.onBoolean flag for whether or not the Coordinator should submit kill task for unused segments, that is, permanently delete them from metadata store and deep storage. If set to true, then for all whitelisted datasources (or optionally all), Coordinator will submit tasks periodically based on period specified. A whitelist can be set via dynamic configuration killDataSourceWhitelist described later.

When druid.coordinator.kill.on is true, segments are eligible for permanent deletion once their data intervals are older than druid.coordinator.kill.durationToRetain relative to the current time. If a segment's data interval is older than this threshold at the time it is marked unused, it is eligible for permanent deletion immediately after being marked unused.
false
druid.coordinator.kill.periodThe frequency of sending kill tasks to the indexing service. The value must be greater than or equal to druid.coordinator.period.indexingPeriod. Only applies if kill is turned on.Same as druid.coordinator.period.indexingPeriod
druid.coordinator.kill.durationToRetainOnly applies if you set druid.coordinator.kill.on to true. This value is ignored if druid.coordinator.kill.ignoreDurationToRetain is true. Valid configurations must be a ISO8601 period. Druid will not kill unused segments whose interval end date is beyond now - durationToRetain. durationToRetain can be a negative ISO8601 period, which would result in now - durationToRetain to be in the future.

Note that the durationToRetain parameter applies to the segment interval, not the time that the segment was last marked unused. For example, if durationToRetain is set to P90D, then a segment for a time chunk 90 days in the past is eligible for permanent deletion immediately after being marked unused.
P90D
druid.coordinator.kill.ignoreDurationToRetainA way to override druid.coordinator.kill.durationToRetain and tell the coordinator that you do not care about the end date of unused segment intervals when it comes to killing them. If true, the coordinator considers all unused segments as eligible to be killed.false
druid.coordinator.kill.bufferPeriodThe amount of time that a segment must be unused before it is able to be permanently removed from metadata and deep storage. This can serve as a buffer period to prevent data loss if data ends up being needed after being marked unused.P30D
druid.coordinator.kill.maxSegmentsThe number of unused segments to kill per kill task. This number must be greater than 0. This only applies when druid.coordinator.kill.on=true.100
druid.coordinator.balancer.strategySpecify the type of balancing strategy for the Coordinator to use to distribute segments among the Historical services. diskNormalized weights the costs according to the servers' disk usage ratios - there are known issues with this strategy distributing segments unevenly across the cluster. random distributes segments among services randomly.cost
druid.coordinator.loadqueuepeon.http.repeatDelayThe start and repeat delay (in milliseconds) for the load queue peon, which manages the load/drop queue of segments for any server.1 minute
druid.coordinator.loadqueuepeon.http.batchSizeNumber of segment load/drop requests to batch in one HTTP request. Note that it must be smaller than druid.segmentCache.numLoadingThreads config on Historical service.1
druid.coordinator.asOverlord.enabledBoolean value for whether this Coordinator service should act like an Overlord as well. This configuration allows users to simplify a Druid cluster by not having to deploy any standalone Overlord services. If set to true, then Overlord console is available at http://coordinator-host:port/console.html and be sure to set druid.coordinator.asOverlord.overlordService also.false
druid.coordinator.asOverlord.overlordServiceRequired, if druid.coordinator.asOverlord.enabled is true. This must be same value as druid.service on standalone Overlord services and druid.selectors.indexing.serviceName on Middle Managers.NULL
Metadata management
PropertyDescriptionRequiredDefault
druid.coordinator.period.metadataStoreManagementPeriodHow often to run metadata management tasks in ISO 8601 duration format.NoPT1H
druid.coordinator.kill.supervisor.onBoolean value for whether to enable automatic deletion of terminated supervisors. If set to true, Coordinator will periodically remove terminated supervisors from the supervisor table in metadata storage.Notrue
druid.coordinator.kill.supervisor.periodHow often to do automatic deletion of terminated supervisor in ISO 8601 duration format. Value must be equal to or greater than druid.coordinator.period.metadataStoreManagementPeriod. Only applies if druid.coordinator.kill.supervisor.on is set to true.NoP1D
druid.coordinator.kill.supervisor.durationToRetainDuration of terminated supervisor to be retained from created time in ISO 8601 duration format. Only applies if druid.coordinator.kill.supervisor.on is set to true.Yes if druid.coordinator.kill.supervisor.on is set to true.P90D
druid.coordinator.kill.audit.onBoolean value for whether to enable automatic deletion of audit logs. If set to true, Coordinator will periodically remove audit logs from the audit table entries in metadata storage.NoTrue
druid.coordinator.kill.audit.periodHow often to do automatic deletion of audit logs in ISO 8601 duration format. Value must be equal to or greater than druid.coordinator.period.metadataStoreManagementPeriod. Only applies if druid.coordinator.kill.audit.on is set to true.NoP1D
druid.coordinator.kill.audit.durationToRetainDuration of audit logs to be retained from created time in ISO 8601 duration format. Only applies if druid.coordinator.kill.audit.on is set to true.Yes if druid.coordinator.kill.audit.on is set to true.P90D
druid.coordinator.kill.compaction.onBoolean value for whether to enable automatic deletion of compaction configurations. If set to true, Coordinator will periodically remove compaction configuration of inactive datasource (datasource with no used and unused segments) from the config table in metadata storage.NoFalse
druid.coordinator.kill.compaction.periodHow often to do automatic deletion of compaction configurations in ISO 8601 duration format. Value must be equal to or greater than druid.coordinator.period.metadataStoreManagementPeriod. Only applies if druid.coordinator.kill.compaction.on is set to true.NoP1D
druid.coordinator.kill.rule.onBoolean value for whether to enable automatic deletion of rules. If set to true, Coordinator will periodically remove rules of inactive datasource (datasource with no used and unused segments) from the rule table in metadata storage.NoTrue
druid.coordinator.kill.rule.periodHow often to do automatic deletion of rules in ISO 8601 duration format. Value must be equal to or greater than druid.coordinator.period.metadataStoreManagementPeriod. Only applies if druid.coordinator.kill.rule.on is set to true.NoP1D
druid.coordinator.kill.rule.durationToRetainDuration of rules to be retained from created time in ISO 8601 duration format. Only applies if druid.coordinator.kill.rule.on is set to true.Yes if druid.coordinator.kill.rule.on is set to true.P90D
druid.coordinator.kill.datasource.onBoolean value for whether to enable automatic deletion of datasource metadata (Note: datasource metadata only exists for datasource created from supervisor). If set to true, Coordinator will periodically remove datasource metadata of terminated supervisor from the datasource table in metadata storage.NoTrue
druid.coordinator.kill.datasource.periodHow often to do automatic deletion of datasource metadata in ISO 8601 duration format. Value must be equal to or greater than druid.coordinator.period.metadataStoreManagementPeriod. Only applies if druid.coordinator.kill.datasource.on is set to true.NoP1D
druid.coordinator.kill.datasource.durationToRetainDuration of datasource metadata to be retained from created time in ISO 8601 duration format. Only applies if druid.coordinator.kill.datasource.on is set to true.Yes if druid.coordinator.kill.datasource.on is set to true.P90D
druid.coordinator.kill.segmentSchema.onBoolean value for whether to enable automatic deletion of unused segment schemas. If set to true, Coordinator will periodically identify segment schemas which are not referenced by any used segment and mark them as unused. At a later point, these unused schemas are deleted. Only applies if Centralized Datasource schema feature is enabled.NoTrue
druid.coordinator.kill.segmentSchema.periodHow often to do automatic deletion of segment schemas in ISO 8601 duration format. Value must be equal to or greater than druid.coordinator.period.metadataStoreManagementPeriod. Only applies if druid.coordinator.kill.segmentSchema.on is set to true.NoP1D
druid.coordinator.kill.segmentSchema.durationToRetainDuration of segment schemas to be retained from the time it was marked as unused in ISO 8601 duration format. Only applies if druid.coordinator.kill.segmentSchema.on is set to true.Yes, if druid.coordinator.kill.segmentSchema.on is set to true.P90D
Segment management
PropertyPossible valuesDescriptionDefault
druid.serverview.typebatch or httpSegment discovery method to use. "http" enables discovering segments using HTTP instead of ZooKeeper.http
druid.coordinator.segment.awaitInitializationOnStarttrue or falseWhether the Coordinator will wait for its view of segments to fully initialize before starting up. If set to 'true', the Coordinator's HTTP server will not start up, and the Coordinator will not announce itself as available, until the server view is initialized.true
Metadata retrieval
PropertyDescriptionDefault
druid.manager.config.pollDurationHow often the manager polls the config table for updates.PT1M
druid.manager.segments.pollDurationThe duration between polls the Coordinator does for updates to the set of active segments. Generally defines the amount of lag time it can take for the Coordinator to notice new segments.PT1M
druid.manager.rules.pollDurationThe duration between polls the Coordinator does for updates to the set of active rules. Generally defines the amount of lag time it can take for the Coordinator to notice rules.PT1M
druid.manager.rules.defaultRuleThe default rule for the cluster_default
druid.manager.rules.alertThresholdThe duration after a failed poll upon which an alert should be emitted.PT10M

Dynamic configuration

The Coordinator has dynamic configurations to tune certain behavior on the fly, without requiring a service restart. You can configure these parameters using the web console(recommended) or through the Coordinator dynamic configuration API.

The following table shows the dynamic configuration properties for the Coordinator.

PropertyDescriptionDefault
millisToWaitBeforeDeletingHow long does the Coordinator need to be a leader before it can start marking overshadowed segments as unused in metadata storage.900000 (15 mins)
mergeBytesLimitThe maximum total uncompressed size in bytes of segments to merge.524288000L
mergeSegmentsLimitThe maximum number of segments that can be in a single append task.100
smartSegmentLoadingEnables "smart" segment loading mode which dynamically computes the optimal values of several properties that maximize Coordinator performance.true
maxSegmentsToMoveThe maximum number of segments that can be moved in a Historical tier at any given time.100
replicantLifetimeThe maximum number of Coordinator runs for which a segment can wait in the load queue of a Historical before Druid raises an alert.15
replicationThrottleLimitThe maximum number of segment replicas that can be assigned to a historical tier in a single Coordinator run. This property prevents Historical services from becoming overwhelmed when loading extra replicas of segments that are already available in the cluster.500
balancerComputeThreadsThread pool size for computing moving cost of segments during segment balancing. Consider increasing this if you have a lot of segments and moving segments begins to stall.num_cores / 2
killDataSourceWhitelistList of specific data sources for which kill tasks are sent if property druid.coordinator.kill.on is true. This can be a list of comma-separated data source names or a JSON array.none
killTaskSlotRatioRatio of total available task slots, including autoscaling if applicable that will be allowed for kill tasks. This value must be between 0 and 1. Only applicable for kill tasks that are spawned automatically by the coordinator's auto kill duty, which is enabled when druid.coordinator.kill.on is true.0.1
maxKillTaskSlotsMaximum number of tasks that will be allowed for kill tasks. This limit only applies for kill tasks that are spawned automatically by the coordinator's auto kill duty, which is enabled when druid.coordinator.kill.on is true.Integer.MAX_VALUE - no limit
killPendingSegmentsSkipListList of data sources for which pendingSegments are NOT cleaned up if property druid.coordinator.kill.pendingSegments.on is true. This can be a list of comma-separated data sources or a JSON array.none
maxSegmentsInNodeLoadingQueueThe maximum number of segments allowed in the load queue of any given server. Use this parameter to load segments faster if, for example, the cluster contains slow-loading nodes or if there are too many segments to be replicated to a particular node (when faster loading is preferred to better segments distribution). The optimal value depends on the loading speed of segments, acceptable replication time and number of nodes.500
useRoundRobinSegmentAssignmentBoolean flag for whether segments should be assigned to Historical services in a round robin fashion. When disabled, segment assignment is done using the chosen balancer strategy. When enabled, this can speed up segment assignments leaving balancing to move the segments to their optimal locations (based on the balancer strategy) lazily.true
decommissioningNodesList of Historical servers to decommission. Coordinator will not assign new segments to decommissioning servers, and segments will be moved away from them to be placed on non-decommissioning servers at the maximum rate specified by maxSegmentsToMove.none
pauseCoordinationBoolean flag for whether or not the Coordinator should execute its various duties of coordinating the cluster. Setting this to true essentially pauses all coordination work while allowing the API to remain up. Duties that are paused include all classes that implement the CoordinatorDuty interface. Such duties include: segment balancing, segment compaction, submitting kill tasks for unused segments (if enabled), logging of used segments in the cluster, marking of newly unused or overshadowed segments, matching and execution of load/drop rules for used segments, unloading segments that are no longer marked as used from Historical servers. An example of when an admin may want to pause coordination would be if they are doing deep storage maintenance on HDFS name nodes with downtime and don't want the Coordinator to be directing Historical nodes to hit the name node with API requests until maintenance is done and the deep store is declared healthy for use again.false
replicateAfterLoadTimeoutBoolean flag for whether or not additional replication is needed for segments that have failed to load due to the expiry of druid.coordinator.load.timeout. If this is set to true, the Coordinator will attempt to replicate the failed segment on a different historical server. This helps improve the segment availability if there are a few slow Historicals in the cluster. However, the slow Historical may still load the segment later and the Coordinator may issue drop requests if the segment is over-replicated.false
Smart segment loading

The smartSegmentLoading mode simplifies Coordinator configuration for segment loading and balancing. If you enable this mode, do not provide values for the properties in the table below as the Coordinator computes them automatically. Druid computes the values to optimize Coordinator performance, based on the current state of the cluster.

If you enable smartSegmentLoading mode, Druid ignores any value you provide for the following properties.

PropertyComputed valueDescription
useRoundRobinSegmentAssignmenttrueSpeeds up segment assignment.
maxSegmentsInNodeLoadingQueue0Removes the limit on load queue size.
replicationThrottleLimit5% of used segments, minimum value 100Prevents aggressive replication when a Historical disappears only intermittently.
replicantLifetime60Allows segments to wait about an hour (assuming a Coordinator period of 1 minute) in the load queue before an alert is raised. In smartSegmentLoading mode, load queues are not limited by size. Segments might therefore assigned to a load queue even if the corresponding server is slow to load them.
maxSegmentsToMove2% of used segments, minimum value 100, maximum value 1000Ensures that some segments are always moving in the cluster to keep it well balanced. The maximum value keeps the Coordinator run times bounded.
balancerComputeThreadsnum_cores / 2Ensures that there are enough threads to perform balancing computations without hogging all Coordinator resources.

When smartSegmentLoading is disabled, Druid uses the configured values of these properties. Disable smartSegmentLoading only if you want to explicitly set the values of any of the above properties.

Lookups dynamic configuration

These configuration options control Coordinator lookup management. For configurations that affect lookup propagation, see Dynamic configuration for lookups.

PropertyDescriptionDefault
druid.manager.lookups.hostDeleteTimeoutHow long to wait for a DELETE request to a particular service before considering the DELETE a failure.PT1S
druid.manager.lookups.hostUpdateTimeoutHow long to wait for a POST request to a particular service before considering the POST a failure.PT10S
druid.manager.lookups.deleteAllTimeoutHow long to wait for all DELETE requests to finish before considering the delete attempt a failure.PT10S
druid.manager.lookups.updateAllTimeoutHow long to wait for all POST requests to finish before considering the attempt a failure.PT60S
druid.manager.lookups.threadPoolSizeHow many services can be managed concurrently (concurrent POST and DELETE requests). Requests this limit will wait in a queue until a slot becomes available.10
druid.manager.lookups.periodNumber of milliseconds between checks for configuration changes.120000 (2 minutes)
Automatic compaction dynamic configuration

You can set or update automatic compaction properties dynamically using the Automatic compaction API without restarting Coordinators.

For details about segment compaction, see Segment size optimization.

You can configure automatic compaction through the following properties:

PropertyDescriptionRequired
dataSourceThe datasource name to be compacted.yes
taskPriorityPriority of compaction task.no (default = 25)
inputSegmentSizeBytesMaximum number of total segment bytes processed per compaction task. Since a time chunk must be processed in its entirety, if the segments for a particular time chunk have a total size in bytes greater than this parameter, compaction will not run for that time chunk.no (default = 100,000,000,000,000 i.e. 100TB)
skipOffsetFromLatestThe offset for searching segments to be compacted in ISO 8601 duration format. Strongly recommended to set for realtime datasources. See Data handling with compaction.no (default = "P1D")
tuningConfigTuning config for compaction tasks. See below Automatic compaction tuningConfig.no
taskContextTask context for compaction tasks.no
granularitySpecCustom granularitySpec. See Automatic compaction granularitySpec.no
dimensionsSpecCustom dimensionsSpec. See Automatic compaction dimensionsSpec.no
transformSpecCustom transformSpec. See Automatic compaction transformSpec.no
metricsSpecCustom metricsSpec. The compaction task preserves any existing metrics regardless of whether metricsSpec is specified. If metricsSpec is specified, Druid does not reapply any aggregators matching the metric names specified in metricsSpec to rows that already have the associated metrics. For rows that do not already have the metric specified in metricsSpec, Druid applies the metric aggregator on the source column, then proceeds to combine the metrics across segments as usual. If metricsSpec is not specified, Druid automatically discovers the metrics in the existing segments and combines existing metrics with the same metric name across segments. Aggregators for metrics with the same name are assumed to be compatible for combining across segments, otherwise the compaction task may fail.no
ioConfigIO config for compaction tasks. See Automatic compaction ioConfig.no

Automatic compaction config example:

{
"dataSource": "wikiticker",
"granularitySpec" : {
"segmentGranularity" : "none"
}
}

Compaction tasks fail when higher priority tasks cause Druid to revoke their locks. By default, realtime tasks like ingestion have a higher priority than compaction tasks. Frequent conflicts between compaction tasks and realtime tasks can cause the Coordinator's automatic compaction to hang. You may see this issue with streaming ingestion from Kafka and Kinesis, which ingest late-arriving data.

To mitigate this problem, set skipOffsetFromLatest to a value large enough so that arriving data tends to fall outside the offset value from the current time. This way you can avoid conflicts between compaction tasks and realtime ingestion tasks. For example, if you want to skip over segments from thirty days prior to the end time of the most recent segment, assign "skipOffsetFromLatest": "P30D". For more information, see Avoid conflicts with ingestion.

Automatic compaction tuningConfig

Auto-compaction supports a subset of the tuningConfig for Parallel task.

The following table shows the supported configurations for auto-compaction.

PropertyDescriptionRequired
typeThe task type. If you're using Coordinator duties for auto-compaction, set it to index_parallel. If you're using compaction supervisors, set it to autocompact.yes
maxRowsInMemoryUsed in determining when intermediate persists to disk should occur. Normally user does not need to set this, but depending on the nature of data, if rows are short in terms of bytes, user may not want to store a million rows in memory and this value should be set.no (default = 1000000)
maxBytesInMemoryUsed in determining when intermediate persists to disk should occur. Normally this is computed internally and user does not need to set it. This value represents number of bytes to aggregate in heap memory before persisting. This is based on a rough estimate of memory usage and not actual usage. The maximum heap memory usage for indexing is maxBytesInMemory * (2 + maxPendingPersists)no (default = 1/6 of max JVM memory)
splitHintSpecUsed to give a hint to control the amount of data that each first phase task reads. This hint could be ignored depending on the implementation of the input source. See Split hint spec for more details.no (default = size-based split hint spec)
partitionsSpecDefines how to partition data in each time chunk, see PartitionsSpecno (default = dynamic)
indexSpecDefines segment storage format options to be used at indexing time, see IndexSpecno
indexSpecForIntermediatePersistsDefines segment storage format options to be used at indexing time for intermediate persisted temporary segments. this can be used to disable dimension/metric compression on intermediate segments to reduce memory required for final merging. however, disabling compression on intermediate segments might increase page cache use while they are used before getting merged into final segment published, see IndexSpec for possible values.no
maxPendingPersistsMaximum number of persists that can be pending but not started. If this limit would be exceeded by a new intermediate persist, ingestion will block until the currently-running persist finishes. Maximum heap memory usage for indexing scales with maxRowsInMemory * (2 + maxPendingPersists).no (default = 0, meaning one persist can be running concurrently with ingestion, and none can be queued up)
pushTimeoutMilliseconds to wait for pushing segments. It must be >= 0, where 0 means to wait forever.no (default = 0)
segmentWriteOutMediumFactorySegment write-out medium to use when creating segments. See SegmentWriteOutMediumFactory.no (default is the value from druid.peon.defaultSegmentWriteOutMediumFactory.type is used)
maxNumConcurrentSubTasksMaximum number of worker tasks which can be run in parallel at the same time. The supervisor task would spawn worker tasks up to maxNumConcurrentSubTasks regardless of the current available task slots. If this value is set to 1, the Supervisor task processes data ingestion on its own instead of spawning worker tasks. If this value is set to too large, too many worker tasks can be created which might block other ingestion. Check Capacity Planning for more details.no (default = 1)
maxRetryMaximum number of retries on task failures.no (default = 3)
maxNumSegmentsToMergeMax limit for the number of segments that a single task can merge at the same time in the second phase. Used only with hashed or single_dim partitionsSpec.no (default = 100)
totalNumMergeTasksTotal number of tasks to merge segments in the merge phase when partitionsSpec is set to hashed or single_dim.no (default = 10)
taskStatusCheckPeriodMsPolling period in milliseconds to check running task statuses.no (default = 1000)
chatHandlerTimeoutTimeout for reporting the pushed segments in worker tasks.no (default = PT10S)
chatHandlerNumRetriesRetries for reporting the pushed segments in worker tasks.no (default = 5)
engineEngine for compaction. Can be either native or msq. msq uses the MSQ task engine and is only supported with compaction supervisors.no (default = native)
Automatic compaction granularitySpec
FieldDescriptionRequired
segmentGranularityTime chunking period for the segment granularity. Defaults to 'null', which preserves the original segment granularity. Accepts all Query granularity values.No
queryGranularityThe resolution of timestamp storage within each segment. Defaults to 'null', which preserves the original query granularity. Accepts all Query granularity values.No
rollupWhether to enable ingestion-time rollup or not. Defaults to null, which preserves the original setting. Note that once data is rollup, individual records can no longer be recovered.No
Automatic compaction dimensionsSpec
FieldDescriptionRequired
dimensionsA list of dimension names or objects. Defaults to null, which preserves the original dimensions. Note that setting this will cause segments manually compacted with dimensionExclusions to be compacted again.No
Automatic compaction transformSpec
FieldDescriptionRequired
filterConditionally filters input rows during compaction. Only rows that pass the filter will be included in the compacted segments. Any of Druid's standard query filters can be used. Defaults to null, which will not filter any row.No
Automatic compaction ioConfig

Auto-compaction supports a subset of the ioConfig for Parallel task. The below is a list of the supported configurations for auto-compaction.

PropertyDescriptionDefaultRequired
dropExistingIf true the compaction task replaces all existing segments fully contained by the umbrella interval of the compacted segments when the task publishes new segments and tombstones. If compaction fails, Druid does not publish any segments or tombstones. WARNING: this functionality is still in beta. Note that changing this config does not cause intervals to be compacted again.falseno

Overlord

For general Overlord service information, see Overlord.

Overlord static configuration

These Overlord static configurations can be defined in the overlord/runtime.properties file.

Overlord service configs
PropertyDescriptionDefault
druid.hostThe host for the current service. This is used to advertise the current service location as reachable from another service and should generally be specified such that http://${druid.host}/ could actually talk to this service.InetAddress.getLocalHost().getCanonicalHostName()
druid.bindOnHostIndicating whether the service's internal jetty server bind on druid.host. Default is false, which means binding to all interfaces.false
druid.plaintextPortThis is the port to actually listen on; unless port mapping is used, this will be the same port as is on druid.host.8090
druid.tlsPortTLS port for HTTPS connector, if druid.enableTlsPort is set then this config will be used. If druid.host contains port then that port will be ignored. This should be a non-negative Integer.8290
druid.serviceThe name of the service. This is used as a dimension when emitting metrics and alerts to differentiate between the various services.druid/overlord
Overlord operations
PropertyDescriptionDefault
druid.indexer.runner.typeIndicates whether tasks should be run locally using local or in a distributed environment using remote. The recommended option is httpRemote, which is similar to remote but uses HTTP to interact with Middle Managers instead of ZooKeeper.httpRemote
druid.indexer.storage.typeIndicates whether incoming tasks should be stored locally (in heap) or in metadata storage. One of local or metadata. local is mainly for internal testing while metadata is recommended in production because storing incoming tasks in metadata storage allows for tasks to be resumed if the Overlord should fail.local
druid.indexer.storage.recentlyFinishedThresholdDuration of time to store task results. Default is 24 hours. If you have hundreds of tasks running in a day, consider increasing this threshold.PT24H
druid.indexer.tasklock.forceTimeChunkLockSetting this to false is still experimental
If set, all tasks are enforced to use time chunk lock. If not set, each task automatically chooses a lock type to use. This configuration can be overwritten by setting forceTimeChunkLock in the task context. See Task Locking & Priority for more details about locking in tasks.
true
druid.indexer.tasklock.batchSegmentAllocationIf set to true, Druid performs segment allocate actions in batches to improve throughput and reduce the average task/action/run/time. See batching segmentAllocate actions for details.true
druid.indexer.tasklock.batchAllocationWaitTimeNumber of milliseconds after Druid adds the first segment allocate action to a batch, until it executes the batch. Allows the batch to add more requests and improve the average segment allocation run time. This configuration takes effect only if batchSegmentAllocation is enabled.0
druid.indexer.task.default.contextDefault task context that is applied to all tasks submitted to the Overlord. Any default in this config does not override neither the context values the user provides nor druid.indexer.tasklock.forceTimeChunkLock.empty context
druid.indexer.queue.maxSizeMaximum number of active tasks at one time.Integer.MAX_VALUE
druid.indexer.queue.startDelaySleep this long before starting Overlord queue management. This can be useful to give a cluster time to re-orient itself (for example, after a widespread network issue).PT1M
druid.indexer.queue.restartDelaySleep this long when Overlord queue management throws an exception before trying again.PT30S
druid.indexer.queue.storageSyncRateSync Overlord state this often with an underlying task persistence mechanism.PT1M
druid.indexer.queue.maxTaskPayloadSizeMaximum allowed size in bytes of a single task payload accepted by the Overlord.none (allow all task payload sizes)

The following configs only apply if the Overlord is running in remote mode. For a description of local vs. remote mode, see Overlord service.

PropertyDescriptionDefault
druid.indexer.runner.taskAssignmentTimeoutHow long to wait after a task has been assigned to a Middle Manager before throwing an error.PT5M
druid.indexer.runner.minWorkerVersionThe minimum Middle Manager version to send tasks to. The version number is a string. This affects the expected behavior during certain operations like comparison against druid.worker.version. Specifically, the version comparison follows dictionary order. Use ISO8601 date format for the version to accommodate date comparisons."0"
druid.indexer.runner.parallelIndexTaskSlotRatioThe ratio of task slots available for parallel indexing supervisor tasks per worker. The specified value must be in the range [0, 1].1
druid.indexer.runner.compressZnodesIndicates whether or not the Overlord should expect Middle Managers to compress Znodes.true
druid.indexer.runner.maxZnodeBytesThe maximum size Znode in bytes that can be created in ZooKeeper, should be in the range of [10KiB, 2GiB). Human-readable format is supported.512 KiB
druid.indexer.runner.taskCleanupTimeoutHow long to wait before failing a task after a Middle Manager is disconnected from ZooKeeper.PT15M
druid.indexer.runner.taskShutdownLinkTimeoutHow long to wait on a shutdown request to a Middle Manager before timing outPT1M
druid.indexer.runner.pendingTasksRunnerNumThreadsNumber of threads to allocate pending-tasks to workers, must be at least 1.1
druid.indexer.runner.maxRetriesBeforeBlacklistNumber of consecutive times the Middle Manager can fail tasks, before the worker is blacklisted, must be at least 15
druid.indexer.runner.workerBlackListBackoffTimeHow long to wait before a task is whitelisted again. This value should be greater that the value set for taskBlackListCleanupPeriod.PT15M
druid.indexer.runner.workerBlackListCleanupPeriodA duration after which the cleanup thread will start up to clean blacklisted workers.PT5M
druid.indexer.runner.maxPercentageBlacklistWorkersThe maximum percentage of workers to blacklist, this must be between 0 and 100.20

If autoscaling is enabled, you can set these additional configs:

PropertyDescriptionDefault
druid.indexer.autoscale.strategySets the strategy to run when autoscaling is required. One of noop, ec2 or gce.noop
druid.indexer.autoscale.doAutoscaleIf set to true, autoscaling will be enabled.false
druid.indexer.autoscale.provisionPeriodHow often to check whether or not new Middle Managers should be added.PT1M
druid.indexer.autoscale.terminatePeriodHow often to check when Middle Managers should be removed.PT5M
druid.indexer.autoscale.originTimeThe starting reference timestamp that the terminate period increments upon.2012-01-01T00:55:00.000Z
druid.indexer.autoscale.workerIdleTimeoutHow long can a worker be idle (not a run task) before it can be considered for termination.PT90M
druid.indexer.autoscale.maxScalingDurationHow long the Overlord will wait around for a Middle Manager to show up before giving up.PT15M
druid.indexer.autoscale.numEventsToTrackThe number of autoscaling related events (node creation and termination) to track.10
druid.indexer.autoscale.pendingTaskTimeoutHow long a task can be in "pending" state before the Overlord tries to scale up.PT30S
druid.indexer.autoscale.workerVersionIf set, will only create nodes of set version during autoscaling. Overrides dynamic configuration.null
druid.indexer.autoscale.workerPortThe port that Middle Managers will run on.8080
druid.indexer.autoscale.workerCapacityHintAn estimation of the number of task slots available for each worker launched by the auto scaler when there are no workers running. The auto scaler uses the worker capacity hint to launch workers with an adequate capacity to handle pending tasks. When unset or set to a value less than or equal to 0, the auto scaler scales workers equal to the value for minNumWorkers in autoScaler config instead. The auto scaler assumes that each worker, either a Middle Manager or indexer, has the same amount of task slots. Therefore, when all your workers have the same capacity (homogeneous capacity), set the value for autoscale.workerCapacityHint equal to druid.worker.capacity. If your workers have different capacities (heterogeneous capacity), set the value to the average of druid.worker.capacity across the workers. For example, if two workers have druid.worker.capacity=10, and one has druid.worker.capacity=4, set autoscale.workerCapacityHint=8. Only applies to pendingTaskBased provisioning strategy.-1
Supervisors
PropertyDescriptionDefault
druid.supervisor.healthinessThresholdThe number of successful runs before an unhealthy supervisor is again considered healthy.3
druid.supervisor.unhealthinessThresholdThe number of failed runs before the supervisor is considered unhealthy.3
druid.supervisor.taskHealthinessThresholdThe number of consecutive task successes before an unhealthy supervisor is again considered healthy.3
druid.supervisor.taskUnhealthinessThresholdThe number of consecutive task failures before the supervisor is considered unhealthy.3
druid.supervisor.storeStackTraceWhether full stack traces of supervisor exceptions should be stored and returned by the supervisor /status endpoint.false
druid.supervisor.maxStoredExceptionEventsThe maximum number of exception events that can be returned through the supervisor /status endpoint.max(healthinessThreshold, unhealthinessThreshold)
druid.supervisor.idleConfig.enabledIf true, supervisor can become idle if there is no data on input stream/topic for some time.false
druid.supervisor.idleConfig.inactiveAfterMillisSupervisor is marked as idle if all existing data has been read from input topic and no new data has been published for inactiveAfterMillis milliseconds.600_000

The druid.supervisor.idleConfig.* specification in the Overlord runtime properties defines the default behavior for the entire cluster. See Idle Configuration in Kafka Supervisor IOConfig to override it for an individual supervisor.

Overlord dynamic configuration

The Overlord has dynamic configurations to tune how Druid assigns tasks to workers. You can configure these parameters using the web console or through the Overlord dynamic configuration API.

The following table shows the dynamic configuration properties for the Overlord.

PropertyDescriptionDefault
selectStrategyDescribes how to assign tasks to Middle Managers. The type can be equalDistribution, equalDistributionWithCategorySpec, fillCapacity, fillCapacityWithCategorySpec, and javascript.{"type":"equalDistribution"}
autoScalerOnly used if autoscaling is enabled.null

The following is an example of an Overlord dynamic config:

Click to view the example
{
"selectStrategy": {
"type": "fillCapacity",
"affinityConfig": {
"affinity": {
"datasource1": ["host1:port", "host2:port"],
"datasource2": ["host3:port"]
}
}
},
"autoScaler": {
"type": "ec2",
"minNumWorkers": 2,
"maxNumWorkers": 12,
"envConfig": {
"availabilityZone": "us-east-1a",
"nodeData": {
"amiId": "${AMI}",
"instanceType": "c3.8xlarge",
"minInstances": 1,
"maxInstances": 1,
"securityGroupIds": ["${IDs}"],
"keyName": "${KEY_NAME}"
},
"userData": {
"impl": "string",
"data": "${SCRIPT_COMMAND}",
"versionReplacementString": ":VERSION:",
"version": null
}
}
}
}
Worker select strategy

The select strategy controls how Druid assigns tasks to workers (Middle Managers). At a high level, the select strategy determines the list of eligible workers for a given task using either an affinityConfig or a categorySpec. Then, Druid assigns the task by either trying to distribute load equally (equalDistribution) or to fill as many workers as possible to capacity (fillCapacity). There are 4 options for select strategies:

A javascript option is also available but should only be used for prototyping new strategies.

If an affinityConfig is provided (as part of fillCapacity and equalDistribution strategies) for a given task, the list of workers eligible to be assigned is determined as follows:

  • a non-affinity worker if no affinity is specified for that datasource. Any worker not listed in the affinityConfig is considered a non-affinity worker.
  • a non-affinity worker if preferred workers are not available and the affinity is weak i.e. strong: false.
  • a preferred worker listed in the affinityConfig for this datasource if it has available capacity
  • no worker if preferred workers are not available and affinity is strong i.e. strong: true. In this case, the task remains in "pending" state. The chosen provisioning strategy (e.g. pendingTaskBased) may then use the total number of pending tasks to determine if a new node should be provisioned.

Note that every worker listed in the affinityConfig will only be used for the assigned datasources and no other.

If a categorySpec is provided (as part of fillCapacityWithCategorySpec and equalDistributionWithCategorySpec strategies), then a task of a given datasource may be assigned to:

  • any worker if no category config is given for task type
  • any worker if category config is given for task type but no category is given for datasource and there's no default category
  • a preferred worker (based on category config and category for datasource) if available
  • any worker if category config and category are given but no preferred worker is available and category config is weak
  • not assigned at all if preferred workers are not available and category config is strong

In both the cases, Druid determines the list of eligible workers and selects one depending on their load with the goal of either distributing the load equally or filling as few workers as possible.

If you are using auto-scaling, use the fillCapacity select strategy since auto-scaled nodes can not be assigned a category, and you want the work to be concentrated on the fewest number of workers to allow the empty ones to scale down.

equalDistribution

Tasks are assigned to the Middle Manager with the most free slots at the time the task begins running. This evenly distributes work across your Middle Managers.

PropertyDescriptionDefault
typeequalDistributionrequired; must be equalDistribution
affinityConfigAffinityConfig objectnull (no affinity)
equalDistributionWithCategorySpec

This strategy is a variant of equalDistribution, which supports workerCategorySpec field rather than affinityConfig. By specifying workerCategorySpec, you can assign tasks to run on different categories of Middle Managers based on the type and dataSource of the task. This strategy doesn't work with AutoScaler since the behavior is undefined.

PropertyDescriptionDefault
typeequalDistributionWithCategorySpecrequired; must be equalDistributionWithCategorySpec
workerCategorySpecWorkerCategorySpec objectnull (no worker category spec)

The following example shows tasks of type index_kafka that default to running on Middle Managers of category c1, except for tasks that write to datasource ds1, which run on Middle Managers of category c2.

{
"selectStrategy": {
"type": "equalDistributionWithCategorySpec",
"workerCategorySpec": {
"strong": false,
"categoryMap": {
"index_kafka": {
"defaultCategory": "c1",
"categoryAffinity": {
"ds1": "c2"
}
}
}
}
}
}
fillCapacity

Tasks are assigned to the worker with the most currently-running tasks. This is useful when you are auto-scaling Middle Managers since it tends to pack some full and leave others empty. The empty ones can be safely terminated.

Note that if druid.indexer.runner.pendingTasksRunnerNumThreads is set to N > 1, then this strategy will fill N Middle Managers up to capacity simultaneously, rather than a single Middle Manager.

PropertyDescriptionDefault
typefillCapacityrequired; must be fillCapacity
affinityConfigAffinityConfig objectnull (no affinity)
fillCapacityWithCategorySpec

This strategy is a variant of fillCapacity, which supports workerCategorySpec instead of an affinityConfig. The usage is the same as equalDistributionWithCategorySpec strategy. This strategy doesn't work with AutoScaler since the behavior is undefined.

PropertyDescriptionDefault
typefillCapacityWithCategorySpec.required; must be fillCapacityWithCategorySpec
workerCategorySpecWorkerCategorySpec objectnull (no worker category spec)
javascript

Allows defining arbitrary logic for selecting workers to run task using a JavaScript function. The function is passed remoteTaskRunnerConfig, map of workerId to available workers and task to be executed and returns the workerId on which the task should be run or null if the task cannot be run. It can be used for rapid development of missing features where the worker selection logic is to be changed or tuned often. If the selection logic is quite complex and cannot be easily tested in JavaScript environment, its better to write a druid extension module with extending current worker selection strategies written in java.

PropertyDescriptionDefault
typejavascriptrequired; must be javascript
functionString representing JavaScript function

The following example shows a function that sends batch_index_task to workers 10.0.0.1 and 10.0.0.2 and all other tasks to other available workers.

{
"type":"javascript",
"function":"function (config, zkWorkers, task) {\nvar batch_workers = new java.util.ArrayList();\nbatch_workers.add(\"middleManager1_hostname:8091\");\nbatch_workers.add(\"middleManager2_hostname:8091\");\nworkers = zkWorkers.keySet().toArray();\nvar sortedWorkers = new Array()\n;for(var i = 0; i < workers.length; i++){\n sortedWorkers[i] = workers[i];\n}\nArray.prototype.sort.call(sortedWorkers,function(a, b){return zkWorkers.get(b).getCurrCapacityUsed() - zkWorkers.get(a).getCurrCapacityUsed();});\nvar minWorkerVer = config.getMinWorkerVersion();\nfor (var i = 0; i < sortedWorkers.length; i++) {\n var worker = sortedWorkers[i];\n var zkWorker = zkWorkers.get(worker);\n if(zkWorker.canRunTask(task) && zkWorker.isValidVersion(minWorkerVer)){\n if(task.getType() == 'index_hadoop' && batch_workers.contains(worker)){\n return worker;\n } else {\n if(task.getType() != 'index_hadoop' && !batch_workers.contains(worker)){\n return worker;\n }\n }\n }\n}\nreturn null;\n}"
}
info

JavaScript-based functionality is disabled by default. Refer to the Druid JavaScript programming guide for guidelines about using Druid's JavaScript functionality, including instructions on how to enable it.

affinityConfig

Use the affinityConfig field to pass affinity configuration to the equalDistribution and fillCapacity strategies. If not provided, the default is to have no affinity.

PropertyDescriptionDefault
affinityJSON object mapping a datasource String name to a list of indexing service Middle Manager host:port values. Druid doesn't perform DNS resolution, so the 'host' value must match what is configured on the Middle Manager and what the Middle Manager announces itself as (examine the Overlord logs to see what your Middle Manager announces itself as).{}
strongWhen true tasks for a datasource must be assigned to affinity-mapped Middle Managers. Tasks remain queued until a slot becomes available. When false, Druid may assign tasks for a datasource to other Middle Managers when affinity-mapped Middle Managers are unavailable to run queued tasks.false
workerCategorySpec

You can provide workerCategorySpec to the equalDistributionWithCategorySpec and fillCapacityWithCategorySpec strategies using the workerCategorySpec field. If not provided, the default is to not use it at all.

PropertyDescriptionDefault
categoryMapA JSON map object mapping a task type String name to a CategoryConfig object, by which you can specify category config for different task type.{}
strongWith weak workerCategorySpec (the default), tasks for a dataSource may be assigned to other Middle Managers if the Middle Managers specified in categoryMap are not able to run all pending tasks in the queue for that dataSource. With strong workerCategorySpec, tasks for a dataSource will only ever be assigned to their specified Middle Managers, and will wait in the pending queue if necessary.false
CategoryConfig
PropertyDescriptionDefault
defaultCategorySpecify default category for a task type.null
categoryAffinityA JSON map object mapping a datasource String name to a category String name of the Middle Manager. If category isn't specified for a datasource, then using the defaultCategory. If no specified category and the defaultCategory is also null, then tasks can run on any available Middle Managers.null
Autoscaler

Amazon's EC2 together with Google's GCE are currently the only supported autoscalers.

EC2's autoscaler properties are:

PropertyDescriptionDefault
typeec20
minNumWorkersThe minimum number of workers that can be in the cluster at any given time.0
maxNumWorkersThe maximum number of workers that can be in the cluster at any given time.0
envConfig.availabilityZoneWhat Amazon availability zone to run in.none
envConfig.nodeDataA JSON object that describes how to launch new nodes.none; required
envConfig.userDataA JSON object that describes how to configure new nodes. If you have set druid.indexer.autoscale.workerVersion, this must have a versionReplacementString. Otherwise, a versionReplacementString is not necessary.none; optional

For GCE's properties, please refer to the gce-extensions.

Data server

This section contains the configuration options for the services that reside on Data servers (Middle Managers/Peons and Historicals) in the suggested three-server configuration.

Configuration options for the Indexer process are also provided here.

Middle Manager and Peon

These Middle Manager and Peon configurations can be defined in the middleManager/runtime.properties file.

Middle Manager service config

PropertyDescriptionDefault
druid.hostThe host for the current service. This is used to advertise the current service location as reachable from another service and should generally be specified such that http://${druid.host}/ could actually talk to this serviceInetAddress.getLocalHost().getCanonicalHostName()
druid.bindOnHostIndicating whether the service's internal jetty server bind on druid.host. Default is false, which means binding to all interfaces.false
druid.plaintextPortThis is the port to actually listen on; unless port mapping is used, this will be the same port as is on druid.host8091
druid.tlsPortTLS port for HTTPS connector, if druid.enableTlsPort is set then this config will be used. If druid.host contains port then that port will be ignored. This should be a non-negative Integer.8291
druid.serviceThe name of the service. This is used as a dimension when emitting metrics and alerts to differentiate between the various servicesdruid/middlemanager

Middle Manager configuration

Middle Managers pass their configurations down to their child peons. The Middle Manager requires the following configs:

PropertyDescriptionDefault
druid.indexer.runner.allowedPrefixesWhitelist of prefixes for configs that can be passed down to child peons.com.metamx, druid, org.apache.druid, user.timezone, file.encoding, java.io.tmpdir, hadoop
druid.indexer.runner.compressZnodesIndicates whether or not the Middle Managers should compress Znodes.true
druid.indexer.runner.classpathJava classpath for the peon.System.getProperty("java.class.path")
druid.indexer.runner.javaCommandCommand required to execute java.java
druid.indexer.runner.javaOptsDEPRECATED A string of -X Java options to pass to the peon's JVM. Quotable parameters or parameters with spaces are encouraged to use javaOptsArray''
druid.indexer.runner.javaOptsArrayA JSON array of strings to be passed in as options to the peon's JVM. This is additive to druid.indexer.runner.javaOpts and is recommended for properly handling arguments which contain quotes or spaces like ["-XX:OnOutOfMemoryError=kill -9 %p"][]
druid.indexer.runner.maxZnodeBytesThe maximum size Znode in bytes that can be created in ZooKeeper, should be in the range of [10KiB, 2GiB). Human-readable format is supported.512KiB
druid.indexer.runner.startPortStarting port used for Peon services, should be greater than 1023 and less than 65536.8100
druid.indexer.runner.endPortEnding port used for Peon services, should be greater than or equal to druid.indexer.runner.startPort and less than 65536.65535
druid.indexer.runner.portsA JSON array of integers to specify ports that used for Peon services. If provided and non-empty, ports for Peon services will be chosen from these ports. And druid.indexer.runner.startPort/druid.indexer.runner.endPort will be completely ignored.[]
druid.worker.ipThe IP of the worker.localhost
druid.worker.versionVersion identifier for the Middle Manager. The version number is a string. This affects the expected behavior during certain operations like comparison against druid.indexer.runner.minWorkerVersion. Specifically, the version comparison follows dictionary order. Use ISO8601 date format for the version to accommodate date comparisons.0
druid.worker.capacityMaximum number of tasks the Middle Manager can accept.Number of CPUs on the machine - 1
druid.worker.baseTaskDirsList of base temporary working directories, one of which is assigned per task in a round-robin fashion. This property can be used to allow usage of multiple disks for indexing. This property is recommended in place of and takes precedence over ${druid.indexer.task.baseTaskDir}. If this configuration is not set, ${druid.indexer.task.baseTaskDir} is used. For example, druid.worker.baseTaskDirs=[\"PATH1\",\"PATH2\",...].null
druid.worker.baseTaskDirSizeThe total amount of bytes that can be used by tasks on any single task dir. This value is treated symmetrically across all directories, that is, if this is 500 GB and there are 3 baseTaskDirs, then each of those task directories is assumed to allow for 500 GB to be used and a total of 1.5 TB will potentially be available across all tasks. The actual amount of memory assigned to each task is discussed in Configuring task storage sizesLong.MAX_VALUE
druid.worker.categoryA string to name the category that the Middle Manager node belongs to._default_worker_category
druid.indexer.fork.property.druid.centralizedDatasourceSchema.enabledThis config should be set when Centralized Datasource Schema feature is enabled.false

Peon processing

Processing properties set on the Middle Manager are passed through to Peons.

PropertyDescriptionDefault
druid.processing.buffer.sizeBytesThis specifies a buffer size (less than 2GiB) for the storage of intermediate results. The computation engine in both the Historical and Realtime processes will use a scratch buffer of this size to do all of their intermediate computations off-heap. Larger values allow for more aggregations in a single pass over the data while smaller values can require more passes depending on the query that is being executed. Human-readable format is supported.auto (max 1 GiB)
druid.processing.buffer.poolCacheMaxCountProcessing buffer pool caches the buffers for later use. This is the maximum count that the cache will grow to. Note that pool can create more buffers than it can cache if necessary.Integer.MAX_VALUE
druid.processing.formatStringRealtime and Historical processes use this format string to name their processing threads.processing-%s
druid.processing.numMergeBuffersThe number of direct memory buffers available for merging query results. The buffers are sized by druid.processing.buffer.sizeBytes. This property is effectively a concurrency limit for queries that require merging buffers. If you are using any queries that require merge buffers (currently, just groupBy) then you should have at least two of these.max(2, druid.processing.numThreads / 4)
druid.processing.numThreadsThe number of processing threads to have available for parallel processing of segments. Our rule of thumb is num_cores - 1, which means that even under heavy load there will still be one core available to do background tasks like talking with ZooKeeper and pulling down segments. If only one core is available, this property defaults to the value 1.Number of cores - 1 (or 1)
druid.processing.fifoEnables the processing queue to treat tasks of equal priority in a FIFO manner.true
druid.processing.tmpDirPath where temporary files created while processing a query should be stored. If specified, this configuration takes priority over the default java.io.tmpdir path.path represented by java.io.tmpdir
druid.processing.intermediaryData.storage.typeStorage type for intermediary segments of data shuffle between native parallel index tasks.
Set to local to store segment files in the local storage of the Middle Manager or Indexer.
Set to deepstore to use configured deep storage for better fault tolerance during rolling updates. When the storage type is deepstore, Druid stores the data in the shuffle-data directory under the configured deep storage path. Druid does not support automated cleanup for the shuffle-data directory. You can set up cloud storage lifecycle rules for automated cleanup of data at the shuffle-data prefix location.
local

The amount of direct memory needed by Druid is at least druid.processing.buffer.sizeBytes * (druid.processing.numMergeBuffers + druid.processing.numThreads + 1). You can ensure at least this amount of direct memory is available by providing -XX:MaxDirectMemorySize=<VALUE> in druid.indexer.runner.javaOptsArray as documented above.

Peon query configuration

See general query configuration.

Peon caching

You can optionally configure caching to be enabled on the peons by setting caching configs here.

PropertyPossible ValuesDescriptionDefault
druid.realtime.cache.useCachetrue, falseEnable the cache on the realtime.false
druid.realtime.cache.populateCachetrue, falsePopulate the cache on the realtime.false
druid.realtime.cache.unCacheableAll druid query typesAll query types to not cache.[]
druid.realtime.cache.maxEntrySizepositive integerMaximum cache entry size in bytes.1_000_000

See cache configuration for how to configure cache settings.

Additional Peon configuration

Although Peons inherit the configurations of their parent Middle Managers, explicit child Peon configs in Middle Manager can be set by prefixing them with:

druid.indexer.fork.property

Additional Peon configs include:

PropertyDescriptionDefault
druid.peon.modeOne of local or remote. Setting this property to local means you intend to run the Peon as a standalone process which is not recommended.remote
druid.indexer.task.baseDirBase temporary working directory.System.getProperty("java.io.tmpdir")
druid.indexer.task.baseTaskDirBase temporary working directory for tasks.${druid.indexer.task.baseDir}/persistent/task
druid.indexer.task.defaultHadoopCoordinatesHadoop version to use with HadoopIndexTasks that do not request a particular version.org.apache.hadoop:hadoop-client-api:3.3.6, org.apache.hadoop:hadoop-client-runtime:3.3.6
druid.indexer.task.defaultRowFlushBoundaryHighest row count before persisting to disk. Used for indexing generating tasks.75000
druid.indexer.task.directoryLockTimeoutWait this long for zombie Peons to exit before giving up on their replacements.PT10M
druid.indexer.task.gracefulShutdownTimeoutWait this long on Middle Manager restart for restorable tasks to gracefully exit.PT5M
druid.indexer.task.hadoopWorkingPathTemporary working directory for Hadoop tasks./tmp/druid-indexing
druid.indexer.task.restoreTasksOnRestartIf true, Middle Managers will attempt to stop tasks gracefully on shutdown and restore them on restart.false
druid.indexer.task.ignoreTimestampSpecForDruidInputSourceIf true, tasks using the Druid input source will ignore the provided timestampSpec, and will use the __time column of the input datasource. This option is provided for compatibility with ingestion specs written before Druid 0.22.0.false
druid.indexer.task.storeEmptyColumnsBoolean value for whether or not to store empty columns during ingestion. When set to true, Druid stores every column specified in the dimensionsSpec. If you use the string-based schemaless ingestion and don't specify any dimensions to ingest, you must also set includeAllDimensions for Druid to store empty columns.

If you set storeEmptyColumns to false, Druid SQL queries referencing empty columns will fail. If you intend to leave storeEmptyColumns disabled, you should either ingest placeholder data for empty columns or else not query on empty columns.

You can overwrite this configuration by setting storeEmptyColumns in the task context.
true
druid.indexer.task.tmpStorageBytesPerTaskMaximum number of bytes per task to be used to store temporary files on disk. This config is generally intended for internal usage. Attempts to set it are very likely to be overwritten by the TaskRunner that executes the task, so be sure of what you expect to happen before directly adjusting this configuration parameter. The config is documented here primarily to provide an understanding of what it means if/when someone sees that it has been set. A value of -1 disables this limit.-1
druid.indexer.server.maxChatRequestsMaximum number of concurrent requests served by a task's chat handler. Set to 0 to disable limiting.0

If the Peon is running in remote mode, there must be an Overlord up and running. Peons in remote mode can set the following configurations:

PropertyDescriptionDefault
druid.peon.taskActionClient.retry.minWaitThe minimum retry time to communicate with Overlord.PT5S
druid.peon.taskActionClient.retry.maxWaitThe maximum retry time to communicate with Overlord.PT1M
druid.peon.taskActionClient.retry.maxRetryCountThe maximum number of retries to communicate with Overlord.60
SegmentWriteOutMediumFactory

When new segments are created, Druid temporarily stores some preprocessed data in some buffers. The following types of medium exist for the buffers:

  • Temporary files (tmpFile) are stored under the task working directory (see druid.worker.baseTaskDirs configuration above) and thus share it's mounting properties. For example, they could be backed by HDD, SSD or memory (tmpfs). This type of medium may do unnecessary disk I/O and requires some disk space to be available.

  • Off-heap memory (offHeapMemory) creates buffers in off-heap memory of a JVM process that is running a task. This type of medium is preferred, but it may require you to allow the JVM to have more off-heap memory by changing the -XX:MaxDirectMemorySize configuration. It's not understood yet how the required off-heap memory size relates to the size of the segments being created. But you shouldn't add more extra off-heap memory than the configured maximum heap size (-Xmx) for the same JVM.

  • On-heap memory (onHeapMemory) creates buffers using the allocated heap memory of the JVM process running a task. Using on-heap memory introduces garbage collection overhead and so is not recommended in most cases. This type of medium is most helpful for tasks run on external clusters where it may be difficult to allocate and work with direct memory effectively.

For most types of tasks, SegmentWriteOutMediumFactory can be configured per-task (see Tasks for more information), but if it's not specified for a task, or it's not supported for a particular task type, then Druid uses the value from the following configuration:

PropertyDescriptionDefault
druid.peon.defaultSegmentWriteOutMediumFactory.typetmpFile, offHeapMemory, or onHeapMemorytmpFile

Indexer

Indexer process configuration

PropertyDescriptionDefault
druid.hostThe host for the current process. This is used to advertise the current processes location as reachable from another process and should generally be specified such that http://${druid.host}/ could actually talk to this processInetAddress.getLocalHost().getCanonicalHostName()
druid.bindOnHostIndicating whether the process's internal jetty server bind on druid.host. Default is false, which means binding to all interfaces.false
druid.plaintextPortThis is the port to actually listen on; unless port mapping is used, this will be the same port as is on druid.host8091
druid.tlsPortTLS port for HTTPS connector, if druid.enableTlsPort is set then this config will be used. If druid.host contains port then that port will be ignored. This should be a non-negative Integer.8283
druid.serviceThe name of the service. This is used as a dimension when emitting metrics and alerts to differentiate between the various servicesdruid/indexer

Indexer general configuration

PropertyDescriptionDefault
druid.worker.versionVersion identifier for the Indexer.0
druid.worker.capacityMaximum number of tasks the Indexer can accept.Number of available processors - 1
druid.worker.baseTaskDirsList of base temporary working directories, one of which is assigned per task in a round-robin fashion. This property can be used to allow usage of multiple disks for indexing. This property is recommended in place of and takes precedence over ${druid.indexer.task.baseTaskDir}. If this configuration is not set, ${druid.indexer.task.baseTaskDir} is used. Example: druid.worker.baseTaskDirs=[\"PATH1\",\"PATH2\",...].null
druid.worker.baseTaskDirSizeThe total amount of bytes that can be used by tasks on any single task dir. This value is treated symmetrically across all directories, that is, if this is 500 GB and there are 3 baseTaskDirs, then each of those task directories is assumed to allow for 500 GB to be used and a total of 1.5 TB will potentially be available across all tasks. The actual amount of memory assigned to each task is discussed in Configuring task storage sizesLong.MAX_VALUE
druid.worker.globalIngestionHeapLimitBytesTotal amount of heap available for ingestion processing. This is applied by automatically setting the maxBytesInMemory property on tasks.Configured max JVM heap size / 6
druid.worker.numConcurrentMergesMaximum number of segment persist or merge operations that can run concurrently across all tasks.druid.worker.capacity / 2, rounded down
druid.indexer.task.baseDirBase temporary working directory.System.getProperty("java.io.tmpdir")
druid.indexer.task.baseTaskDirBase temporary working directory for tasks.${druid.indexer.task.baseDir}/persistent/tasks
druid.indexer.task.defaultHadoopCoordinatesHadoop version to use with HadoopIndexTasks that do not request a particular version.org.apache.hadoop:hadoop-client-api:3.3.6, org.apache.hadoop:hadoop-client-runtime:3.3.6
druid.indexer.task.gracefulShutdownTimeoutWait this long on Indexer restart for restorable tasks to gracefully exit.PT5M
druid.indexer.task.hadoopWorkingPathTemporary working directory for Hadoop tasks./tmp/druid-indexing
druid.indexer.task.restoreTasksOnRestartIf true, the Indexer will attempt to stop tasks gracefully on shutdown and restore them on restart.false
druid.indexer.task.ignoreTimestampSpecForDruidInputSourceIf true, tasks using the Druid input source will ignore the provided timestampSpec, and will use the __time column of the input datasource. This option is provided for compatibility with ingestion specs written before Druid 0.22.0.false
druid.indexer.task.storeEmptyColumnsBoolean value for whether or not to store empty columns during ingestion. When set to true, Druid stores every column specified in the dimensionsSpec.

If you set storeEmptyColumns to false, Druid SQL queries referencing empty columns will fail. If you intend to leave storeEmptyColumns disabled, you should either ingest placeholder data for empty columns or else not query on empty columns.

You can overwrite this configuration by setting storeEmptyColumns in the task context.
true
druid.peon.taskActionClient.retry.minWaitThe minimum retry time to communicate with Overlord.PT5S
druid.peon.taskActionClient.retry.maxWaitThe maximum retry time to communicate with Overlord.PT1M
druid.peon.taskActionClient.retry.maxRetryCountThe maximum number of retries to communicate with Overlord.60

Indexer concurrent requests

Druid uses Jetty to serve HTTP requests.

PropertyDescriptionDefault
druid.server.http.numThreadsNumber of threads for HTTP requests. Please see the Indexer Server HTTP threads documentation for more details on how the Indexer uses this configuration.max(10, (Number of cores * 17) / 16 + 2) + 30
druid.server.http.queueSizeSize of the worker queue used by Jetty server to temporarily store incoming client connections. If this value is set and a request is rejected by jetty because queue is full then client would observe request failure with TCP connection being closed immediately with a completely empty response from server.Unbounded
druid.server.http.maxIdleTimeThe Jetty max idle time for a connection.PT5M
druid.server.http.enableRequestLimitIf enabled, no requests would be queued in jetty queue and "HTTP 429 Too Many Requests" error response would be sent.false
druid.server.http.defaultQueryTimeoutQuery timeout in millis, beyond which unfinished queries will be cancelled300000
druid.server.http.gracefulShutdownTimeoutThe maximum amount of time Jetty waits after receiving shutdown signal. After this timeout the threads will be forcefully shutdown. This allows any queries that are executing to complete(Only values greater than zero are valid).PT30S
druid.server.http.unannouncePropagationDelayHow long to wait for ZooKeeper unannouncements to propagate before shutting down Jetty. This is a minimum and druid.server.http.gracefulShutdownTimeout does not start counting down until after this period elapses.PT0S (do not wait)
druid.server.http.maxQueryTimeoutMaximum allowed value (in milliseconds) for timeout parameter. See query-context to know more about timeout. Query is rejected if the query context timeout is greater than this value.Long.MAX_VALUE
druid.server.http.maxRequestHeaderSizeMaximum size of a request header in bytes. Larger headers consume more memory and can make a server more vulnerable to denial of service attacks.8 * 1024
druid.server.http.enableForwardedRequestCustomizerIf enabled, adds Jetty ForwardedRequestCustomizer which reads X-Forwarded-* request headers to manipulate servlet request object when Druid is used behind a proxy.false
druid.server.http.allowedHttpMethodsList of HTTP methods that should be allowed in addition to the ones required by Druid APIs. Druid APIs require GET, PUT, POST, and DELETE, which are always allowed. This option is not useful unless you have installed an extension that needs these additional HTTP methods or that adds functionality related to CORS. None of Druid's bundled extensions require these methods.[]
druid.server.http.contentSecurityPolicyContent-Security-Policy header value to set on each non-POST response. Setting this property to an empty string, or omitting it, both result in the default frame-ancestors: none being set.frame-ancestors 'none'

Indexer processing resources

PropertyDescriptionDefault
druid.processing.buffer.sizeBytesThis specifies a buffer size (less than 2GiB) for the storage of intermediate results. The computation engine in the Indexer processes will use a scratch buffer of this size to do all of their intermediate computations off-heap. Larger values allow for more aggregations in a single pass over the data while smaller values can require more passes depending on the query that is being executed. Human-readable format is supported.auto (max 1GiB)
druid.processing.buffer.poolCacheMaxCountprocessing buffer pool caches the buffers for later use, this is the maximum count cache will grow to. note that pool can create more buffers than it can cache if necessary.Integer.MAX_VALUE
druid.processing.formatStringIndexer processes use this format string to name their processing threads.processing-%s
druid.processing.numMergeBuffersThe number of direct memory buffers available for merging query results. The buffers are sized by druid.processing.buffer.sizeBytes. This property is effectively a concurrency limit for queries that require merging buffers. If you are using any queries that require merge buffers (currently, just groupBy) then you should have at least two of these.max(2, druid.processing.numThreads / 4)
druid.processing.numThreadsThe number of processing threads to have available for parallel processing of segments. Our rule of thumb is num_cores - 1, which means that even under heavy load there will still be one core available to do background tasks like talking with ZooKeeper and pulling down segments. If only one core is available, this property defaults to the value 1.Number of cores - 1 (or 1)
druid.processing.fifoIf the processing queue should treat tasks of equal priority in a FIFO mannertrue
druid.processing.tmpDirPath where temporary files created while processing a query should be stored. If specified, this configuration takes priority over the default java.io.tmpdir path.path represented by java.io.tmpdir

The amount of direct memory needed by Druid is at least druid.processing.buffer.sizeBytes * (druid.processing.numMergeBuffers + druid.processing.numThreads + 1). You can ensure at least this amount of direct memory is available by providing -XX:MaxDirectMemorySize=<VALUE> at the command line.

Query configurations

See general query configuration.

Indexer caching

You can optionally configure caching to be enabled on the Indexer by setting caching configs here.

PropertyPossible ValuesDescriptionDefault
druid.realtime.cache.useCachetrue, falseEnable the cache on the realtime.false
druid.realtime.cache.populateCachetrue, falsePopulate the cache on the realtime.false
druid.realtime.cache.unCacheableAll druid query typesAll query types to not cache.[]
druid.realtime.cache.maxEntrySizepositive integerMaximum cache entry size in bytes.1_000_000

See cache configuration for how to configure cache settings.

Note that only local caches such as the local-type cache and caffeine cache are supported. If a remote cache such as memcached is used, it will be ignored.

Historical

For general Historical service information, see Historical.

These Historical configurations can be defined in the historical/runtime.properties file.

Historical service configuration

PropertyDescriptionDefault
druid.hostThe host for the current service. This is used to advertise the current service location as reachable from another service and should generally be specified such that http://${druid.host}/ could actually talk to this serviceInetAddress.getLocalHost().getCanonicalHostName()
druid.bindOnHostIndicating whether the service's internal jetty server bind on druid.host. Default is false, which means binding to all interfaces.false
druid.plaintextPortThis is the port to actually listen on; unless port mapping is used, this will be the same port as is on druid.host8083
druid.tlsPortTLS port for HTTPS connector, if druid.enableTlsPort is set then this config will be used. If druid.host contains port then that port will be ignored. This should be a non-negative Integer.8283
druid.serviceThe name of the service. This is used as a dimension when emitting metrics and alerts to differentiate between the various servicesdruid/historical

Historical general configuration

PropertyDescriptionDefault
druid.server.maxSizeThe maximum number of bytes-worth of segments that the service wants assigned to it. The Coordinator service will attempt to assign segments to a Historical service only if this property is greater than the total size of segments served by it. Since this property defines the upper limit on the total segment size that can be assigned to a Historical, it is defaulted to the sum of all maxSize values specified within druid.segmentCache.locations property. Human-readable format is supported, see here.Sum of maxSize values defined within druid.segmentCache.locations
druid.server.tierA string to name the distribution tier that the storage service belongs to. Many of the rules Coordinator services use to manage segments can be keyed on tiers._default_tier
druid.server.priorityIn a tiered architecture, the priority of the tier, thus allowing control over which services are queried. Higher numbers mean higher priority. The default (no priority) works for architecture with no cross replication (tiers that have no data-storage overlap). Data centers typically have equal priority.0

Storing segments

PropertyDescriptionDefault
druid.segmentCache.locationsSegments assigned to a Historical services are first stored on the local file system (in a disk cache) and then served by the Historical services. These locations define where that local cache resides. This value cannot be NULL or EMPTY. Here is an example druid.segmentCache.locations=[{"path": "/mnt/druidSegments", "maxSize": "10k", "freeSpacePercent": 1.0}]. "freeSpacePercent" is optional, if provided then enforces that much of free disk partition space while storing segments. But, it depends on File.getTotalSpace() and File.getFreeSpace() methods, so enable if only if they work for your File System.none
druid.segmentCache.locationSelector.strategyThe strategy used to select a location from the configured druid.segmentCache.locations for segment distribution. Possible values are leastBytesUsed, roundRobin, random, or mostAvailableSize.leastBytesUsed
druid.segmentCache.deleteOnRemoveDelete segment files from cache once a service is no longer serving a segment.true
druid.segmentCache.dropSegmentDelayMillisHow long a service delays before completely dropping segment.30000 (30 seconds)
druid.segmentCache.infoDirHistorical services keep track of the segments they are serving so that when the service is restarted they can reload the same segments without waiting for the Coordinator to reassign. This path defines where this metadata is kept. Directory will be created if needed.${first_location}/info_dir
druid.segmentCache.announceIntervalMillisHow frequently to announce segments while segments are loading from cache. Set this value to zero to wait for all segments to be loaded before announcing.5000 (5 seconds)
druid.segmentCache.numLoadingThreadsHow many segments to drop or load concurrently from deep storage. Note that the work of loading segments involves downloading segments from deep storage, decompressing them and loading them to a memory mapped location. So the work is not all I/O Bound. Depending on CPU and network load, one could possibly increase this config to a higher value.max(1,Number of cores / 6)
druid.segmentCache.numBootstrapThreadsHow many segments to load concurrently during historical startup.druid.segmentCache.numLoadingThreads
druid.segmentCache.lazyLoadOnStartWhether or not to load segment columns metadata lazily during historical startup. When set to true, Historical startup time will be dramatically improved by deferring segment loading until the first time that segment takes part in a query, which will incur this cost instead.false
druid.segmentCache.numThreadsToLoadSegmentsIntoPageCacheOnDownloadNumber of threads to asynchronously read segment index files into null output stream on each new segment download after the Historical service finishes bootstrapping. Recommended to set to 1 or 2 or leave unspecified to disable. See also druid.segmentCache.numThreadsToLoadSegmentsIntoPageCacheOnBootstrap0
druid.segmentCache.numThreadsToLoadSegmentsIntoPageCacheOnBootstrapNumber of threads to asynchronously read segment index files into null output stream during Historical service bootstrap. This thread pool is terminated after Historical service finishes bootstrapping. Recommended to set to half of available cores. If left unspecified, druid.segmentCache.numThreadsToLoadSegmentsIntoPageCacheOnDownload will be used. If both configs are unspecified, this feature is disabled. Preemptively loading segments into page cache helps in the sense that later when a segment is queried, it's already in page cache and only a minor page fault needs to be triggered instead of a more costly major page fault to make the query latency more consistent. Note that loading segment into page cache just does a blind loading of segment index files and will evict any existing segments from page cache at the discretion of operating system when the total segment size on local disk is larger than the page cache usable in the RAM, which roughly equals to total available RAM in the host - druid process memory including both heap and direct memory allocated - memory used by other non druid processes on the host, so it is the user's responsibility to ensure the host has enough RAM to host all the segments to avoid random evictions to fully leverage this feature.druid.segmentCache.numThreadsToLoadSegmentsIntoPageCacheOnDownload

In druid.segmentCache.locations, freeSpacePercent was added because the maxSize setting is only a theoretical limit and assumes that much space will always be available for storing segments. In case of any druid bug leading to unaccounted segment files left alone on disk or some other service writing stuff to disk, This check can start failing segment loading early before filling up the disk completely and leaving the host usable otherwise.

In druid.segmentCache.locationSelector.strategy, one of leastBytesUsed, roundRobin, random, or mostAvailableSize could be specified to represent the strategy to distribute segments across multiple segment cache locations.

StrategyDescription
leastBytesUsedSelects a location which has least bytes used in absolute terms.
roundRobinSelects a location in a round robin fashion oblivious to the bytes used or the capacity.
randomSelects a segment cache location randomly each time among the available storage locations.
mostAvailableSizeSelects a segment cache location that has most free space among the available storage locations.

Note that if druid.segmentCache.numLoadingThreads > 1, multiple threads can download different segments at the same time. In this case, with the leastBytesUsed strategy or mostAvailableSize strategy, Historicals may select a sub-optimal storage location because each decision is based on a snapshot of the storage location status of when a segment is requested to download.

Historical query configs

Concurrent requests

Druid uses Jetty to serve HTTP requests.

PropertyDescriptionDefault
druid.server.http.numThreadsNumber of threads for HTTP requests.max(10, (Number of cores * 17) / 16 + 2) + 30
druid.server.http.queueSizeSize of the worker queue used by Jetty server to temporarily store incoming client connections. If this value is set and a request is rejected by jetty because queue is full then client would observe request failure with TCP connection being closed immediately with a completely empty response from server.Unbounded
druid.server.http.maxIdleTimeThe Jetty max idle time for a connection.PT5M
druid.server.http.enableRequestLimitIf enabled, no requests would be queued in jetty queue and "HTTP 429 Too Many Requests" error response would be sent.false
druid.server.http.defaultQueryTimeoutQuery timeout in millis, beyond which unfinished queries will be cancelled300000
druid.server.http.gracefulShutdownTimeoutThe maximum amount of time Jetty waits after receiving shutdown signal. After this timeout the threads will be forcefully shutdown. This allows any queries that are executing to complete(Only values greater than zero are valid).PT30S
druid.server.http.unannouncePropagationDelayHow long to wait for ZooKeeper unannouncements to propagate before shutting down Jetty. This is a minimum and druid.server.http.gracefulShutdownTimeout does not start counting down until after this period elapses.PT0S (do not wait)
druid.server.http.maxQueryTimeoutMaximum allowed value (in milliseconds) for timeout parameter. See query-context to know more about timeout. Query is rejected if the query context timeout is greater than this value.Long.MAX_VALUE
druid.server.http.maxRequestHeaderSizeMaximum size of a request header in bytes. Larger headers consume more memory and can make a server more vulnerable to denial of service attacks.8 * 1024
druid.server.http.contentSecurityPolicyContent-Security-Policy header value to set on each non-POST response. Setting this property to an empty string, or omitting it, both result in the default frame-ancestors: none being set.frame-ancestors 'none'
Processing
PropertyDescriptionDefault
druid.processing.buffer.sizeBytesThis specifies a buffer size (less than 2GiB), for the storage of intermediate results. The computation engine in both the Historical and Realtime processes will use a scratch buffer of this size to do all of their intermediate computations off-heap. Larger values allow for more aggregations in a single pass over the data while smaller values can require more passes depending on the query that is being executed. Human-readable format is supported.auto (max 1GiB)
druid.processing.buffer.poolCacheMaxCountprocessing buffer pool caches the buffers for later use, this is the maximum count cache will grow to. note that pool can create more buffers than it can cache if necessary.Integer.MAX_VALUE
druid.processing.formatStringRealtime and Historical processes use this format string to name their processing threads.processing-%s
druid.processing.numMergeBuffersThe number of direct memory buffers available for merging query results. The buffers are sized by druid.processing.buffer.sizeBytes. This property is effectively a concurrency limit for queries that require merging buffers. If you are using any queries that require merge buffers (currently, just groupBy) then you should have at least two of these.max(2, druid.processing.numThreads / 4)
druid.processing.numThreadsThe number of processing threads to have available for parallel processing of segments. Our rule of thumb is num_cores - 1, which means that even under heavy load there will still be one core available to do background tasks like talking with ZooKeeper and pulling down segments. If only one core is available, this property defaults to the value 1.Number of cores - 1 (or 1)
druid.processing.fifoIf the processing queue should treat tasks of equal priority in a FIFO mannertrue
druid.processing.tmpDirPath where temporary files created while processing a query should be stored. If specified, this configuration takes priority over the default java.io.tmpdir path.path represented by java.io.tmpdir

The amount of direct memory needed by Druid is at least druid.processing.buffer.sizeBytes * (druid.processing.numMergeBuffers + druid.processing.numThreads + 1). You can ensure at least this amount of direct memory is available by providing -XX:MaxDirectMemorySize=<VALUE> at the command line.

Historical query configuration

See general query configuration.

Historical caching

You can optionally only configure caching to be enabled on the Historical by setting caching configs here.

PropertyPossible ValuesDescriptionDefault
druid.historical.cache.useCachetrue, falseEnable the cache on the Historical.false
druid.historical.cache.populateCachetrue, falsePopulate the cache on the Historical.false
druid.historical.cache.unCacheableAll druid query typesAll query types to not cache.[]
druid.historical.cache.maxEntrySizepositive integerMaximum cache entry size in bytes.1_000_000

See cache configuration for how to configure cache settings.

Query server

This section contains the configuration options for the services that reside on Query servers (Brokers) in the suggested three-server configuration.

Configuration options for the experimental Router process are also provided here.

Broker

For general Broker process information, see here.

These Broker configurations can be defined in the broker/runtime.properties file.

Broker process configs

PropertyDescriptionDefault
druid.hostThe host for the current process. This is used to advertise the current processes location as reachable from another process and should generally be specified such that http://${druid.host}/ could actually talk to this processInetAddress.getLocalHost().getCanonicalHostName()
druid.bindOnHostIndicating whether the process's internal jetty server bind on druid.host. Default is false, which means binding to all interfaces.false
druid.plaintextPortThis is the port to actually listen on; unless port mapping is used, this will be the same port as is on druid.host8082
druid.tlsPortTLS port for HTTPS connector, if druid.enableTlsPort is set then this config will be used. If druid.host contains port then that port will be ignored. This should be a non-negative Integer.8282
druid.serviceThe name of the service. This is used as a dimension when emitting metrics and alerts to differentiate between the various servicesdruid/broker

Query configuration

Query routing
PropertyPossible ValuesDescriptionDefault
druid.broker.balancer.typerandom, connectionCountDetermines how the broker balances connections to Historical processes. random choose randomly, connectionCount picks the process with the fewest number of active connections torandom
druid.broker.select.tierhighestPriority, lowestPriority, customIf segments are cross-replicated across tiers in a cluster, you can tell the broker to prefer to select segments in a tier with a certain priority.highestPriority
druid.broker.select.tier.custom.prioritiesAn array of integer priorities, such as [-1, 0, 1, 2]Select servers in tiers with a custom priority list.The config only has effect if druid.broker.select.tier is set to custom. If druid.broker.select.tier is set to custom but this config is not specified, the effect is the same as druid.broker.select.tier set to highestPriority. Any of the integers in this config can be ignored if there's no corresponding tiers with such priorities. Tiers with priorities explicitly specified in this config always have higher priority than those not and those not specified fall back to use highestPriority strategy among themselves.
Query prioritization and laning

Laning strategies allow you to control capacity utilization for heterogeneous query workloads. With laning, the broker examines and classifies a query for the purpose of assigning it to a lane. Lanes have capacity limits, enforced by the broker, that can be used to ensure sufficient resources are available for other lanes or for interactive queries (with no lane), or to limit overall throughput for queries within the lane. Requests in excess of the capacity are discarded with an HTTP 429 status code.

PropertyDescriptionDefault
druid.query.scheduler.numThreadsMaximum number of concurrently-running queries. When this parameter is set lower than druid.server.http.numThreads, query requests beyond the limit are put into the Jetty request queue. This has the effect of reserving the leftover Jetty threads for non-query requests.

When this parameter is set equal to or higher than druid.server.http.numThreads, it has no effect.
Unbounded
druid.query.scheduler.laning.strategyQuery laning strategy to use to assign queries to a lane in order to control capacities for certain classes of queries.none
druid.query.scheduler.prioritization.strategyQuery prioritization strategy to automatically assign priorities.manual
Prioritization strategies
Manual prioritization strategy

With this configuration, queries are never assigned a priority automatically, but will preserve a priority manually set on the query context with the priority key. This mode can be explicitly set by setting druid.query.scheduler.prioritization.strategy to manual.

Threshold prioritization strategy

This prioritization strategy lowers the priority of queries that cross any of a configurable set of thresholds, such as how far in the past the data is, how large of an interval a query covers, or the number of segments taking part in a query.

This strategy can be enabled by setting druid.query.scheduler.prioritization.strategy to threshold.

PropertyDescriptionDefault
druid.query.scheduler.prioritization.periodThresholdISO duration threshold for how old data can be queried before automatically adjusting query priority.none
druid.query.scheduler.prioritization.durationThresholdISO duration threshold for maximum duration a queries interval can span before the priority is automatically adjusted.none
druid.query.scheduler.prioritization.segmentCountThresholdNumber threshold for maximum number of segments that can take part in a query before its priority is automatically adjusted.none
druid.query.scheduler.prioritization.adjustmentAmount to reduce the priority of queries which cross any threshold.none
Laning strategies
No laning strategy

In this mode, queries are never assigned a lane, and the concurrent query count will only be limited by druid.server.http.numThreads or druid.query.scheduler.numThreads, if set. This is the default Druid query scheduler operating mode. Enable this strategy explicitly by setting druid.query.scheduler.laning.strategy to none.

'High/Low' laning strategy

This laning strategy splits queries with a priority below zero into a low query lane, automatically. Queries with priority of zero (the default) or above are considered 'interactive'. The limit on low queries can be set to some desired percentage of the total capacity (or HTTP thread pool size), reserving capacity for interactive queries. Queries in the low lane are not guaranteed their capacity, which may be consumed by interactive queries, but may use up to this limit if total capacity is available.

If the low lane is specified in the query context lane parameter, this will override the computed lane.

This strategy can be enabled by setting druid.query.scheduler.laning.strategy=hilo.

PropertyDescriptionDefault
druid.query.scheduler.laning.maxLowPercentMaximum percent of the smaller number of druid.server.http.numThreads or druid.query.scheduler.numThreads, defining the number of HTTP threads that can be used by queries with a priority lower than 0. Value must be an integer in the range 1 to 100, and will be rounded upNo default, must be set if using this mode
Guardrails for materialization of subqueries

Druid stores the subquery rows in temporary tables that live in the Java heap. It is a good practice to avoid large subqueries in Druid. Therefore, there are guardrails that are built in Druid to prevent the queries from generating subquery results which can exhaust the heap space. They can be set on a cluster level or modified per query level as desired. Note the following guardrails that can be set by the cluster admin to limit the subquery results:

  1. druid.server.http.maxSubqueryRows in broker's config to set a default for the entire cluster or maxSubqueryRows in the query context to set an upper limit on the number of rows a subquery can generate
  2. druid.server.http.maxSubqueryBytes in broker's config to set a default for the entire cluster or maxSubqueryBytes in the query context to set an upper limit on the number of bytes a subquery can generate

Limiting the subquery by bytes is an experimental feature as it materializes the results differently.

You can configure maxSubqueryBytes to the following values:

  • disabled: It is the default setting out of the box. It disables the subquery's from the byte based limit, and effectively disables this feature.
  • auto: Druid automatically decides the optimal byte based limit based upon the heap space available and the max number of concurrent queries.
  • A positive long value: User can manually specify the number of bytes that the results of the subqueries of a single query can occupy on the heap.

Due to the conversion between the Java objects and the Frame's format, setting maxSubqueryBytes can become slow if the subquery starts generating rows in the order of magnitude of around 10 million and above. In those scenarios, disable the maxSubqueryBytes settings for such queries, assess the number of rows that the subqueries generate and override the maxSubqueryRows to appropriate value.

If you choose to modify or set any of the above limits, you must also think about the heap size of all Brokers, Historicals, and task Peons that process data for the subqueries to accommodate the subquery results. There is no formula to calculate the correct value. Trial and error is the best approach.

Manual laning strategy

This laning strategy is best suited for cases where one or more external applications which query Druid are capable of manually deciding what lane a given query should belong to. Configured with a map of lane names to percent or exact max capacities, queries with a matching lane parameter in the query context will be subjected to those limits.

PropertyDescriptionDefault
druid.query.scheduler.laning.lanes.{name}Maximum percent or exact limit of queries that can concurrently run in the defined lanes. Any number of lanes may be defined like this. The lane names 'total' and 'default' are reserved for internal use.No default, must define at least one lane with a limit above 0. If druid.query.scheduler.laning.isLimitPercent is set to true, values must be integers in the range of 1 to 100.
druid.query.scheduler.laning.isLimitPercentIf set to true, the values set for druid.query.scheduler.laning.lanes will be treated as a percent of the smaller number of druid.server.http.numThreads or druid.query.scheduler.numThreads. Note that in this mode, these lane values across lanes are not required to add up to, and can exceed, 100%.false
Server configuration

Druid uses Jetty to serve HTTP requests. Each query being processed consumes a single thread from druid.server.http.numThreads, so consider defining druid.query.scheduler.numThreads to a lower value in order to reserve HTTP threads for responding to health checks, lookup loading, and other non-query, (in most cases) comparatively very short-lived, HTTP requests.

PropertyDescriptionDefault
druid.server.http.numThreadsNumber of threads for HTTP requests.max(10, (Number of cores * 17) / 16 + 2) + 30
druid.server.http.queueSizeSize of the worker queue used by Jetty server to temporarily store incoming client connections. If this value is set and a request is rejected by jetty because queue is full then client would observe request failure with TCP connection being closed immediately with a completely empty response from server.Unbounded
druid.server.http.maxIdleTimeThe Jetty max idle time for a connection.PT5M
druid.server.http.enableRequestLimitIf enabled, no requests would be queued in jetty queue and "HTTP 429 Too Many Requests" error response would be sent.false
druid.server.http.defaultQueryTimeoutQuery timeout in millis, beyond which unfinished queries will be cancelled300000
druid.server.http.maxScatterGatherBytesMaximum number of bytes gathered from data processes such as Historicals and realtime processes to execute a query. Queries that exceed this limit will fail. This is an advance configuration that allows to protect in case Broker is under heavy load and not utilizing the data gathered in memory fast enough and leading to OOMs. This limit can be further reduced at query time using maxScatterGatherBytes in the context. Note that having large limit is not necessarily bad if broker is never under heavy concurrent load in which case data gathered is processed quickly and freeing up the memory used. Human-readable format is supported, see here.Long.MAX_VALUE
druid.server.http.maxSubqueryRowsMaximum number of rows from all subqueries per query. Druid stores the subquery rows in temporary tables that live in the Java heap. druid.server.http.maxSubqueryRows is a guardrail to prevent the system from exhausting available heap. When a subquery exceeds the row limit, Druid throws a resource limit exceeded exception: "Subquery generated results beyond maximum."

It is a good practice to avoid large subqueries in Druid. However, if you choose to raise the subquery row limit, you must also increase the heap size of all Brokers, Historicals, and task Peons that process data for the subqueries to accommodate the subquery results.

There is no formula to calculate the correct value. Trial and error is the best approach.
100000
druid.server.http.maxSubqueryBytesMaximum number of bytes from all subqueries per query. Since the results are stored on the Java heap, druid.server.http.maxSubqueryBytes is a guardrail like druid.server.http.maxSubqueryRows to prevent the heap space from exhausting. When a subquery exceeds the byte limit, Druid throws a resource limit exceeded exception. A negative value for the guardrail indicates that Druid won't guardrail by memory. This can be set to 'disabled' which disables the results from being limited via the byte limit, 'auto' which sets this value automatically taking free heap space into account, or a positive long value depicting the number of bytes per query's subqueries' results can occupy. This is an experimental feature for now as this materializes the results in a different format.'disabled'
druid.server.http.gracefulShutdownTimeoutThe maximum amount of time Jetty waits after receiving shutdown signal. After this timeout the threads will be forcefully shutdown. This allows any queries that are executing to complete(Only values greater than zero are valid).PT30S
druid.server.http.unannouncePropagationDelayHow long to wait for ZooKeeper unannouncements to propagate before shutting down Jetty. This is a minimum and druid.server.http.gracefulShutdownTimeout does not start counting down until after this period elapses.PT0S (do not wait)
druid.server.http.maxQueryTimeoutMaximum allowed value (in milliseconds) for timeout parameter. See query-context to know more about timeout. Query is rejected if the query context timeout is greater than this value.Long.MAX_VALUE
druid.server.http.maxRequestHeaderSizeMaximum size of a request header in bytes. Larger headers consume more memory and can make a server more vulnerable to denial of service attacks.8 * 1024
druid.server.http.contentSecurityPolicyContent-Security-Policy header value to set on each non-POST response. Setting this property to an empty string, or omitting it, both result in the default frame-ancestors: none being set.frame-ancestors 'none'
druid.server.http.enableHSTSIf set to true, druid services will add strict transport security header Strict-Transport-Security: max-age=63072000; includeSubDomains to all HTTP responsesfalse
Client configuration

Druid Brokers use an HTTP client to communicate with data servers (Historical servers and real-time tasks). This client has the following configuration options.

PropertyDescriptionDefault
druid.broker.http.numConnectionsSize of connection pool for the Broker to connect to Historical and real-time processes. If there are more queries than this number that all need to speak to the same process, then they will queue up.20
druid.broker.http.eagerInitializationIndicates that http connections from Broker to Historical and Real-time processes should be eagerly initialized. If set to true, numConnections connections are created upon initializationtrue
druid.broker.http.compressionCodecCompression codec the Broker uses to communicate with Historical and real-time processes. May be "gzip" or "identity".gzip
druid.broker.http.readTimeoutThe timeout for data reads from Historical servers and real-time tasks.PT15M
druid.broker.http.unusedConnectionTimeoutThe timeout for idle connections in connection pool. The connection in the pool will be closed after this timeout and a new one will be established. This timeout should be less than druid.broker.http.readTimeout. Set this timeout = ~90% of druid.broker.http.readTimeoutPT4M
druid.broker.http.maxQueuedBytesMaximum number of bytes queued per query before exerting backpressure on channels to the data servers.

Similar to druid.server.http.maxScatterGatherBytes, except that maxQueuedBytes triggers backpressure instead of query failure. Set to zero to disable. You can override this setting by using the maxQueuedBytes query context parameter. Druid supports human-readable format.
25 MB or 2% of maximum Broker heap size, whichever is greater.
druid.broker.http.numMaxThreads`Maximum number of I/O worker threadsmax(10, ((number of cores * 17) / 16 + 2) + 30)`
druid.broker.http.clientConnectTimeoutThe timeout (in milliseconds) for establishing client connections.500
Retry policy

Druid broker can optionally retry queries internally for transient errors.

PropertyDescriptionDefault
druid.broker.retryPolicy.numTriesNumber of tries.1
Processing

The broker uses processing configs for nested groupBy queries.

PropertyDescriptionDefault
druid.processing.buffer.sizeBytesThis specifies a buffer size (less than 2GiB) for the storage of intermediate results. The computation engine in both the Historical and Realtime processes will use a scratch buffer of this size to do all of their intermediate computations off-heap. Larger values allow for more aggregations in a single pass over the data while smaller values can require more passes depending on the query that is being executed. Human-readable format is supported.auto (max 1GiB)
druid.processing.buffer.poolCacheInitialCountinitializes the number of buffers allocated on the intermediate results pool. Note that pool can create more buffers if necessary.0
druid.processing.buffer.poolCacheMaxCountprocessing buffer pool caches the buffers for later use, this is the maximum count cache will grow to. note that pool can create more buffers than it can cache if necessary.Integer.MAX_VALUE
druid.processing.numMergeBuffersThe number of direct memory buffers available for merging query results. The buffers are sized by druid.processing.buffer.sizeBytes. This property is effectively a concurrency limit for queries that require merging buffers. If you are using any queries that require merge buffers (currently, just groupBy) then you should have at least two of these.max(2, druid.processing.numThreads / 4)
druid.processing.fifoIf the processing queue should treat tasks of equal priority in a FIFO mannertrue
druid.processing.tmpDirPath where temporary files created while processing a query should be stored. If specified, this configuration takes priority over the default java.io.tmpdir path.path represented by java.io.tmpdir
druid.processing.merge.useParallelMergePoolEnable automatic parallel merging for Brokers on a dedicated async ForkJoinPool. If false, instead merges will be done serially on the HTTP thread pool.true
druid.processing.merge.pool.parallelismSize of ForkJoinPool. Note that the default configuration assumes that the value returned by Runtime.getRuntime().availableProcessors() represents 2 hyper-threads per physical core, and multiplies this value by 0.75 in attempt to size 1.5 times the number of physical cores.Runtime.getRuntime().availableProcessors() * 0.75 (rounded up)
druid.processing.merge.pool.defaultMaxQueryParallelismDefault maximum number of parallel merge tasks per query. Note that the default configuration assumes that the value returned by Runtime.getRuntime().availableProcessors() represents 2 hyper-threads per physical core, and multiplies this value by 0.5 in attempt to size to the number of physical cores.Runtime.getRuntime().availableProcessors() * 0.5 (rounded up)
druid.processing.merge.pool.awaitShutdownMillisTime to wait for merge ForkJoinPool tasks to complete before ungracefully stopping on process shutdown in milliseconds.60_000
druid.processing.merge.task.targetRunTimeMillisIdeal run-time of each ForkJoinPool merge task, before forking off a new task to continue merging sequences.100
druid.processing.merge.task.initialYieldNumRowsNumber of rows to yield per ForkJoinPool merge task, before forking off a new task to continue merging sequences.16384
druid.processing.merge.task.smallBatchNumRowsSize of result batches to operate on in ForkJoinPool merge tasks.4096

The amount of direct memory needed by Druid is at least druid.processing.buffer.sizeBytes * (druid.processing.numMergeBuffers + 1). You can ensure at least this amount of direct memory is available by providing -XX:MaxDirectMemorySize=<VALUE> at the command line.

Broker query configuration

See general query configuration.

Broker generated query configuration supplementation

The Broker generates queries internally. This configuration section describes how an operator can augment the configuration of these queries.

As of now the only supported augmentation is overriding the default query context. This allows an operator the flexibility to adjust it as they see fit. A common use of this configuration is to override the query priority of the cluster generated queries in order to avoid running as a default priority of 0.

PropertyDescriptionDefault
druid.broker.internal.query.config.contextA string formatted key:value map of a query context to add to internally generated broker queries.null

SQL

The Druid SQL server is configured through the following properties on the Broker.

PropertyDescriptionDefault
druid.sql.enableWhether to enable SQL at all, including background metadata fetching. If false, this overrides all other SQL-related properties and disables SQL metadata, serving, and planning completely.true
druid.sql.avatica.enableWhether to enable JDBC querying at /druid/v2/sql/avatica/.true
druid.sql.avatica.maxConnectionsMaximum number of open connections for the Avatica server. These are not HTTP connections, but are logical client connections that may span multiple HTTP connections.25
druid.sql.avatica.maxRowsPerFrameMaximum acceptable value for the JDBC client Statement.setFetchSize method. This setting determines the maximum number of rows that Druid will populate in a single 'fetch' for a JDBC ResultSet. Set this property to -1 to enforce no row limit on the server-side and potentially return the entire set of rows on the initial statement execution. If the JDBC client calls Statement.setFetchSize with a value other than -1, Druid uses the lesser value of the client-provided limit and maxRowsPerFrame. If maxRowsPerFrame is smaller than minRowsPerFrame, then the ResultSet size will be fixed. To handle queries that produce results with a large number of rows, you can increase value of druid.sql.avatica.maxRowsPerFrame to reduce the number of fetches required to completely transfer the result set.5,000
druid.sql.avatica.minRowsPerFrameMinimum acceptable value for the JDBC client Statement.setFetchSize method. The value for this property must greater than 0. If the JDBC client calls Statement.setFetchSize with a lesser value, Druid uses minRowsPerFrame instead. If maxRowsPerFrame is less than minRowsPerFrame, Druid uses the minimum value of the two. For handling queries which produce results with a large number of rows, you can increase this value to reduce the number of fetches required to completely transfer the result set.100
druid.sql.avatica.maxStatementsPerConnectionMaximum number of simultaneous open statements per Avatica client connection.4
druid.sql.avatica.connectionIdleTimeoutAvatica client connection idle timeout.PT5M
druid.sql.avatica.fetchTimeoutMsAvatica fetch timeout, in milliseconds. When a request for the next batch of data takes longer than this time, Druid returns an empty result set, causing the client to poll again. This avoids HTTP timeouts for long-running queries. The default of 5 sec. is good for most cases.5000
druid.sql.http.enableWhether to enable JSON over HTTP querying at /druid/v2/sql/.true
druid.sql.planner.maxTopNLimitMaximum threshold for a TopN query. Higher limits will be planned as GroupBy queries instead.100000
druid.sql.planner.metadataRefreshPeriodThrottle for metadata refreshes.PT1M
druid.sql.planner.metadataColumnTypeMergePolicyDefines how column types will be chosen when faced with differences between segments when computing the SQL schema. Options are specified as a JSON object, with valid choices of leastRestrictive or latestInterval. For leastRestrictive, Druid will automatically widen the type computed for the schema to a type which data across all segments can be converted into, however planned schema migrations can only take effect once all segments have been re-ingested to the new schema. With latestInterval, the column type in most recent time chunks defines the type for the schema.leastRestrictive
druid.sql.planner.useApproximateCountDistinctWhether to use an approximate cardinality algorithm for COUNT(DISTINCT foo).true
druid.sql.planner.useGroupingSetForExactDistinctOnly relevant when useApproximateCountDistinct is disabled. If set to true, exact distinct queries are re-written using grouping sets. Otherwise, exact distinct queries are re-written using joins. This should be set to true for group by query with multiple exact distinct aggregations. This flag can be overridden per query.false
druid.sql.planner.useApproximateTopNWhether to use approximate TopN queries when a SQL query could be expressed as such. If false, exact GroupBy queries will be used instead.true
druid.sql.planner.requireTimeConditionWhether to require SQL to have filter conditions on __time column so that all generated native queries will have user specified intervals. If true, all queries without filter condition on __time column will failfalse
druid.sql.planner.sqlTimeZoneSets the default time zone for the server, which will affect how time functions and timestamp literals behave. Should be a time zone name like "America/Los_Angeles" or offset like "-08:00".UTC
druid.sql.planner.metadataSegmentCacheEnableWhether to keep a cache of published segments in broker. If true, broker polls coordinator in background to get segments from metadata store and maintains a local cache. If false, coordinator's REST API will be invoked when broker needs published segments info.false
druid.sql.planner.metadataSegmentPollPeriodHow often to poll coordinator for published segments list if druid.sql.planner.metadataSegmentCacheEnable is set to true. Poll period is in milliseconds.60000
druid.sql.planner.authorizeSystemTablesDirectlyIf true, Druid authorizes queries against any of the system schema tables (sys in SQL) as SYSTEM_TABLE resources which require READ access, in addition to permissions based content filtering.false
druid.sql.planner.useNativeQueryExplainIf true, EXPLAIN PLAN FOR will return the explain plan as a JSON representation of equivalent native query(s), else it will return the original version of explain plan generated by Calcite. It can be overridden per query with useNativeQueryExplain context key.true
druid.sql.planner.maxNumericInFiltersMax limit for the amount of numeric values that can be compared for a string type dimension when the entire SQL WHERE clause of a query translates to an OR of Bound filter. By default, Druid does not restrict the amount of numeric Bound Filters on String columns, although this situation may block other queries from running. Set this property to a smaller value to prevent Druid from running queries that have prohibitively long segment processing times. The optimal limit requires some trial and error; we recommend starting with 100. Users who submit a query that exceeds the limit of maxNumericInFilters should instead rewrite their queries to use strings in the WHERE clause instead of numbers. For example, WHERE someString IN (‘123’, ‘456’). If this value is disabled, maxNumericInFilters set through query context is ignored.-1 (disabled)
druid.sql.approxCountDistinct.functionImplementation to use for the APPROX_COUNT_DISTINCT function. Without extensions loaded, the only valid value is APPROX_COUNT_DISTINCT_BUILTIN (a HyperLogLog, or HLL, based implementation). If the DataSketches extension is loaded, this can also be APPROX_COUNT_DISTINCT_DS_HLL (alternative HLL implementation) or APPROX_COUNT_DISTINCT_DS_THETA.

Theta sketches use significantly more memory than HLL sketches, so you should prefer one of the two HLL implementations.
APPROX_COUNT_DISTINCT_BUILTIN
info

Previous versions of Druid had properties named druid.sql.planner.maxQueryCount and druid.sql.planner.maxSemiJoinRowsInMemory. These properties are no longer available. Since Druid 0.18.0, you can use druid.server.http.maxSubqueryRows to control the maximum number of rows permitted across all subqueries.

Broker caching

You can optionally only configure caching to be enabled on the Broker by setting caching configs here.

PropertyPossible ValuesDescriptionDefault
druid.broker.cache.useCachetrue, falseEnable the cache on the Broker.false
druid.broker.cache.populateCachetrue, falsePopulate the cache on the Broker.false
druid.broker.cache.useResultLevelCachetrue, falseEnable result level caching on the Broker.false
druid.broker.cache.populateResultLevelCachetrue, falsePopulate the result level cache on the Broker.false
druid.broker.cache.resultLevelCacheLimitpositive integerMaximum size of query response that can be cached.Integer.MAX_VALUE
druid.broker.cache.unCacheableAll druid query typesAll query types to not cache.[]
druid.broker.cache.cacheBulkMergeLimitpositive integer or 0Queries with more segments than this number will not attempt to fetch from cache at the broker level, leaving potential caching fetches (and cache result merging) to the HistoricalsInteger.MAX_VALUE
druid.broker.cache.maxEntrySizepositive integerMaximum cache entry size in bytes.1_000_000

See cache configuration for how to configure cache settings.

info

Note: Even if cache is enabled, for groupBy queries, segment level cache does not work on Brokers. See Query caching for more information.

Segment discovery

PropertyPossible ValuesDescriptionDefault
druid.serverview.typebatch or httpSegment discovery method to use. "http" enables discovering segments using HTTP instead of ZooKeeper.http
druid.broker.segment.watchedTiersList of stringsThe Broker watches segment announcements from processes that serve segments to build a cache to relate each process to the segments it serves. This configuration allows the Broker to only consider segments being served from a list of tiers. By default, Broker considers all tiers. This can be used to partition your dataSources in specific Historical tiers and configure brokers in partitions so that they are only queryable for specific dataSources. This config is mutually exclusive from druid.broker.segment.ignoredTiers and at most one of these can be configured on a Broker.none
druid.broker.segment.ignoredTiersList of stringsThe Broker watches segment announcements from processes that serve segments to build a cache to relate each process to the segments it serves. This configuration allows the Broker to ignore the segments being served from a list of tiers. By default, Broker considers all tiers. This config is mutually exclusive from druid.broker.segment.watchedTiers and at most one of these can be configured on a Broker.none
druid.broker.segment.watchedDataSourcesList of stringsBroker watches the segment announcements from processes serving segments to build cache of which process is serving which segments, this configuration allows to only consider segments being served from a whitelist of dataSources. By default, Broker would consider all datasources. This can be used to configure brokers in partitions so that they are only queryable for specific dataSources.none
druid.broker.segment.watchRealtimeTasksBooleanThe Broker watches segment announcements from processes that serve segments to build a cache to relate each process to the segments it serves. When watchRealtimeTasks is true, the Broker watches for segment announcements from both Historicals and realtime processes. To configure a broker to exclude segments served by realtime processes, set watchRealtimeTasks to false.true
druid.broker.segment.awaitInitializationOnStartBooleanWhether the Broker will wait for its view of segments to fully initialize before starting up. If set to 'true', the Broker's HTTP server will not start up, and the Broker will not announce itself as available, until the server view is initialized. See also druid.sql.planner.awaitInitializationOnStart, a related setting.true

Cache configuration

This section describes caching configuration that is common to Broker, Historical, and Middle Manager/Peon processes.

Caching could optionally be enabled on the Broker, Historical, and Middle Manager/Peon processes. See Broker, Historical, and Peon configuration options for how to enable it for different processes.

Druid uses a local in-memory cache by default, unless a different type of cache is specified. Use the druid.cache.type configuration to set a different kind of cache.

Cache settings are set globally, so the same configuration can be re-used for both Broker and Historical processes, when defined in the common properties file.

Cache type

PropertyPossible ValuesDescriptionDefault
druid.cache.typelocal, memcached, hybrid, caffeineThe type of cache to use for queries. See below of the configuration options for each cache typecaffeine

Local cache

info

DEPRECATED: Use caffeine (default as of v0.12.0) instead

The local cache is deprecated in favor of the Caffeine cache, and may be removed in a future version of Druid. The Caffeine cache affords significantly better performance and control over eviction behavior compared to local cache, and is recommended in any situation where you are using JRE 8u60 or higher.

A simple in-memory LRU cache. Local cache resides in JVM heap memory, so if you enable it, make sure you increase heap size accordingly.

PropertyDescriptionDefault
druid.cache.sizeInBytesMaximum cache size in bytes. Zero disables caching.0
druid.cache.initialSizeInitial size of the hash table backing the cache.500000
druid.cache.logEvictionCountIf non-zero, log cache eviction every logEvictionCount items.0

Caffeine cache

A highly performant local cache implementation for Druid based on Caffeine. Requires a JRE8u60 or higher if using COMMON_FJP.

Configuration

The following table shows the configuration options known to this module:

runtime.propertiesDescriptionDefault
druid.cache.typeSet this to caffeine or leave out parametercaffeine
druid.cache.sizeInBytesThe maximum size of the cache in bytes on heap. It can be configured as described in here.min(1GiB, Runtime.maxMemory / 10)
druid.cache.expireAfterThe time (in ms) after an access for which a cache entry may be expiredNone (no time limit)
druid.cache.cacheExecutorFactoryThe executor factory to use for Caffeine maintenance. One of COMMON_FJP, SINGLE_THREAD, or SAME_THREADForkJoinPool common pool (COMMON_FJP)
druid.cache.evictOnCloseIf a close of a namespace (ex: removing a segment from a process) should cause an eager eviction of associated cache valuesfalse
druid.cache.cacheExecutorFactory

The following are the possible values for druid.cache.cacheExecutorFactory, which controls how maintenance tasks are run:

  • COMMON_FJP (default) use the common ForkJoinPool. Should use with JRE 8u60 or higher. Older versions of the JRE may have worse performance than newer JRE versions.
  • SINGLE_THREAD Use a single-threaded executor.
  • SAME_THREAD Cache maintenance is done eagerly.
Metrics

In addition to the normal cache metrics, the caffeine cache implementation also reports the following in both total and delta:

MetricDescriptionNormal value
query/cache/caffeine/*/requestsCount of hits or misses.hit + miss
query/cache/caffeine/*/loadTimeLength of time caffeine spends loading new values (unused feature).0
query/cache/caffeine/*/evictionBytesSize in bytes that have been evicted from the cacheVaries, should tune cache sizeInBytes so that sizeInBytes/evictionBytes is approximately the rate of cache churn you desire.
Memcached

Uses memcached as cache backend. This allows all processes to share the same cache.

PropertyDescriptionDefault
druid.cache.expirationMemcached expiration time.2592000 (30 days)
druid.cache.timeoutMaximum time in milliseconds to wait for a response from Memcached.500
druid.cache.hostsComma separated list of Memcached hosts <host:port>. Need to specify all nodes when druid.cache.clientMode is set to static. Dynamic mode automatically identifies nodes in your cluster so just specifying the configuration endpoint and port is fine.none
druid.cache.maxObjectSizeMaximum object size in bytes for a Memcached object.52428800 (50 MiB)
druid.cache.memcachedPrefixKey prefix for all keys in Memcached.druid
druid.cache.numConnectionsNumber of memcached connections to use.1
druid.cache.protocolMemcached communication protocol. Can be binary or text.binary
druid.cache.locatorMemcached locator. Can be consistent or array_mod.consistent
druid.cache.enableTlsEnable TLS based connection for Memcached client. Boolean.false
druid.cache.clientModeClient Mode. Static mode requires the user to specify individual cluster nodes. Dynamic mode uses AutoDiscovery feature of AWS Memcached. String. "static" or "dynamic"static
druid.cache.skipTlsHostnameVerificationSkip TLS Hostname Verification. Boolean.true

Hybrid

Uses a combination of any two caches as a two-level L1 / L2 cache. This may be used to combine a local in-memory cache with a remote memcached cache.

Cache requests will first check L1 cache before checking L2. If there is an L1 miss and L2 hit, it will also populate L1.

PropertyDescriptionDefault
druid.cache.l1.typeThe type of cache to use for L1 cache. See druid.cache.type configuration for valid types.caffeine
druid.cache.l2.typeThe type of cache to use for L2 cache. See druid.cache.type configuration for valid types.caffeine
druid.cache.l1.*Any property valid for the given type of L1 cache can be set using this prefix. For instance, if you are using a caffeine L1 cache, specify druid.cache.l1.sizeInBytes to set its size.defaults are the same as for the given cache type
druid.cache.l2.*Prefix for L2 cache settings, see description for L1.defaults are the same as for the given cache type
druid.cache.useL2A boolean indicating whether to query L2 cache, if it's a miss in L1. It makes sense to configure this to false on Historical processes, if L2 is a remote cache like memcached, and this cache also used on brokers, because in this case if a query reached Historical it means that a broker didn't find corresponding results in the same remote cache, so a query to the remote cache from Historical is guaranteed to be a miss.true
druid.cache.populateL2A boolean indicating whether to put results into L2 cache.true

General query configuration

This section describes configurations that control behavior of Druid's query types, applicable to Broker, Historical, and Middle Manager processes.

Overriding default query context values

You can override any query context general parameter default value by setting the runtime property in the format of druid.query.default.context.{query_context_key}. The druid.query.default.context.{query_context_key} runtime property prefix applies to all current and future query context keys, the same as how query context parameter passed with the query works. You can override the runtime property value if the value for the same key is specified in the query contexts.

The precedence chain for query context values is as follows:

hard-coded default value in Druid code <- runtime property not prefixed with druid.query.default.context <- runtime property prefixed with druid.query.default.context <- context parameter in the query

Note that not all query context key has a runtime property not prefixed with druid.query.default.context that can override the hard-coded default value. For example, maxQueuedBytes has druid.broker.http.maxQueuedBytes but joinFilterRewriteMaxSize does not. Hence, the only way of overriding joinFilterRewriteMaxSize hard-coded default value is with runtime property druid.query.default.context.joinFilterRewriteMaxSize.

To further elaborate on the previous example:

If neither druid.broker.http.maxQueuedBytes or druid.query.default.context.maxQueuedBytes is set and the query does not have maxQueuedBytes in the context, then the hard-coded value in Druid code is use. If runtime property only contains druid.broker.http.maxQueuedBytes=x and query does not have maxQueuedBytes in the context, then the value of the property, x, is use. However, if query does have maxQueuedBytes in the context, then that value is use instead. If runtime property only contains druid.query.default.context.maxQueuedBytes=y OR runtime property contains both druid.broker.http.maxQueuedBytes=x and druid.query.default.context.maxQueuedBytes=y, then the value of druid.query.default.context.maxQueuedBytes, y, is use (given that query does not have maxQueuedBytes in the context). If query does have maxQueuedBytes in the context, then that value is use instead.

TopN query config

PropertyDescriptionDefault
druid.query.topN.minTopNThresholdSee TopN Aliasing for details.1000

Search query config

PropertyDescriptionDefault
druid.query.search.maxSearchLimitMaximum number of search results to return.1000
druid.query.search.searchStrategyDefault search query strategy.useIndexes

SegmentMetadata query config

PropertyDescriptionDefault
druid.query.segmentMetadata.defaultHistoryWhen no interval is specified in the query, use a default interval of defaultHistory before the end time of the most recent segment, specified in ISO8601 format. This property also controls the duration of the default interval used by GET /druid/v2/datasources/{dataSourceName} interactions for retrieving datasource dimensions and metrics.P1W
druid.query.segmentMetadata.defaultAnalysisTypesThis can be used to set the Default Analysis Types for all segment metadata queries, this can be overridden when making the query["cardinality", "interval", "minmax"]

GroupBy query config

This section describes the configurations for groupBy queries. You can set the runtime properties in the runtime.properties file on Broker, Historical, and Middle Manager processes. You can set the query context parameters through the query context.

Supported runtime properties:

PropertyDescriptionDefault
druid.query.groupBy.maxSelectorDictionarySizeMaximum amount of heap space (approximately) to use for per-segment string dictionaries. See groupBy memory tuning and resource limits for details.100000000
druid.query.groupBy.maxMergingDictionarySizeMaximum amount of heap space (approximately) to use for per-query string dictionaries. When the dictionary exceeds this size, a spill to disk will be triggered. See groupBy memory tuning and resource limits for details.100000000
druid.query.groupBy.maxOnDiskStorageMaximum amount of disk space to use, per-query, for spilling result sets to disk when either the merging buffer or the dictionary fills up. Queries that exceed this limit will fail. Set to zero to disable disk spilling.0 (disabled)
druid.query.groupBy.defaultOnDiskStorageDefault amount of disk space to use, per-query, for spilling the result sets to disk when either the merging buffer or the dictionary fills up. Set to zero to disable disk spilling for queries which don't override maxOnDiskStorage in their context.druid.query.groupBy.maxOnDiskStorage

Supported query contexts:

KeyDescription
maxSelectorDictionarySizeCan be used to lower the value of druid.query.groupBy.maxMergingDictionarySize for this query.
maxMergingDictionarySizeCan be used to lower the value of druid.query.groupBy.maxMergingDictionarySize for this query.
maxOnDiskStorageCan be used to set maxOnDiskStorage to a value between 0 and druid.query.groupBy.maxOnDiskStorage for this query. If this query context override exceeds druid.query.groupBy.maxOnDiskStorage, the query will use druid.query.groupBy.maxOnDiskStorage. Omitting this from the query context will cause the query to use druid.query.groupBy.defaultOnDiskStorage for maxOnDiskStorage

Advanced configurations

Supported runtime properties:

PropertyDescriptionDefault
druid.query.groupBy.singleThreadedMerge results using a single thread.false
druid.query.groupBy.bufferGrouperInitialBucketsInitial number of buckets in the off-heap hash table used for grouping results. Set to 0 to use a reasonable default (1024).0
druid.query.groupBy.bufferGrouperMaxLoadFactorMaximum load factor of the off-heap hash table used for grouping results. When the load factor exceeds this size, the table will be grown or spilled to disk. Set to 0 to use a reasonable default (0.7).0
druid.query.groupBy.forceHashAggregationForce to use hash-based aggregation.false
druid.query.groupBy.intermediateCombineDegreeNumber of intermediate processes combined together in the combining tree. Higher degrees will need less threads which might be helpful to improve the query performance by reducing the overhead of too many threads if the server has sufficiently powerful CPU cores.8
druid.query.groupBy.numParallelCombineThreadsHint for the number of parallel combining threads. This should be larger than 1 to turn on the parallel combining feature. The actual number of threads used for parallel combining is min(druid.query.groupBy.numParallelCombineThreads, druid.processing.numThreads).1 (disabled)

Supported query contexts:

KeyDescriptionDefault
groupByIsSingleThreadedOverrides the value of druid.query.groupBy.singleThreaded for this query.
bufferGrouperInitialBucketsOverrides the value of druid.query.groupBy.bufferGrouperInitialBuckets for this query.none
bufferGrouperMaxLoadFactorOverrides the value of druid.query.groupBy.bufferGrouperMaxLoadFactor for this query.none
forceHashAggregationOverrides the value of druid.query.groupBy.forceHashAggregationnone
intermediateCombineDegreeOverrides the value of druid.query.groupBy.intermediateCombineDegreenone
numParallelCombineThreadsOverrides the value of druid.query.groupBy.numParallelCombineThreadsnone
sortByDimsFirstSort the results first by dimension values and then by timestamp.false
forceLimitPushDownWhen all fields in the orderby are part of the grouping key, the broker will push limit application down to the Historical processes. When the sorting order uses fields that are not in the grouping key, applying this optimization can result in approximate results with unknown accuracy, so this optimization is disabled by default in that case. Enabling this context flag turns on limit push down for limit/orderbys that contain non-grouping key columns.false

Expression processing configurations

KeyDescriptionDefault
druid.expressions.useStrictBooleansControls the behavior of Druid boolean operators and functions, if set to true all boolean values are either 1 or 0. This configuration has been deprecated and will be removed in a future release, taking on the true behavior. See expression documentation for more information.true

Router

Router process configs

PropertyDescriptionDefault
druid.hostThe host for the current process. This is used to advertise the current processes location as reachable from another process and should generally be specified such that http://${druid.host}/ could actually talk to this processInetAddress.getLocalHost().getCanonicalHostName()
druid.bindOnHostIndicating whether the process's internal jetty server bind on druid.host. Default is false, which means binding to all interfaces.false
druid.plaintextPortThis is the port to actually listen on; unless port mapping is used, this will be the same port as is on druid.host8888
druid.tlsPortTLS port for HTTPS connector, if druid.enableTlsPort is set then this config will be used. If druid.host contains port then that port will be ignored. This should be a non-negative Integer.9088
druid.serviceThe name of the service. This is used as a dimension when emitting metrics and alerts to differentiate between the various servicesdruid/router

Runtime configuration

PropertyDescriptionDefault
druid.router.defaultBrokerServiceNameThe default Broker to connect to in case service discovery fails.druid/broker
druid.router.tierToBrokerMapQueries for a certain tier of data are routed to their appropriate Broker. This value should be an ordered JSON map of tiers to Broker names. The priority of Brokers is based on the ordering.{"_default_tier": "<defaultBrokerServiceName>"}
druid.router.defaultRuleThe default rule for all datasources._default
druid.router.pollPeriodHow often to poll for new rules.PT1M
druid.router.sql.enableEnable routing of SQL queries using strategies. Whentrue, the Router uses the strategies defined in druid.router.strategies to determine the broker service for a given SQL query. When false, the Router uses the defaultBrokerServiceName.false
druid.router.strategiesPlease see Router Strategies for details.[{"type":"timeBoundary"},{"type":"priority"}]
druid.router.avatica.balancer.typeClass to use for balancing Avatica queries across Brokers. Please see Avatica Query Balancing.rendezvousHash
druid.router.managementProxy.enabledEnables the Router's management proxy functionality.false
druid.router.http.numConnectionsSize of connection pool for the Router to connect to Broker processes. If there are more queries than this number that all need to speak to the same process, then they will queue up.20
druid.router.http.eagerInitializationIndicates that http connections from Router to Broker should be eagerly initialized. If set to true, numConnections connections are created upon initializationtrue
druid.router.http.readTimeoutThe timeout for data reads from Broker processes.PT15M
druid.router.http.numMaxThreadsMaximum number of worker threads to handle HTTP requests and responsesmax(10, ((number of cores * 17) / 16 + 2) + 30)
druid.router.http.numRequestsQueuedMaximum number of requests that may be queued to a destination1024
druid.router.http.requestBuffersizeSize of the content buffer for receiving requests. These buffers are only used for active connections that have requests with bodies that will not fit within the header buffer8 * 1024
druid.router.http.clientConnectTimeoutThe timeout (in milliseconds) for establishing client connections.500