Elasticsearch 7.3 CPU Usage

[Spock_Tgai]

Hi,

We are running elasticsearch 7.3 release in 5 node cluster. Two of the nodes in the cluster are suddenly showing constant 50% CPU.

The other nodes are showing less than 20% CPU, any ideas on how to understand why is this ?

[Christian_Dahlqvist]

What does the hot threads API give for those nodes?

[Spock_Tgai]

This is the output I see using the hot threads api.

::: {cbb28bbe97f9}{42Ez3ypVQOCwC355eOjGjw}{ZCw4xARoQ6u6yVHd0SwLAg}{x.x.x.x}{x.x.x.x:9300}{dim}{aws_availability_zone=us-west-2c, ml.machine_memory=16040214528, xpack.installed=true, ml.max_open_jobs=20}
   Hot threads at 2019-08-21T03:09:18.186Z, interval=500ms, busiestThreads=3, ignoreIdleThreads=true:

::: {c91da54d3211}{F6FYAOjQRIq1Oe_00a4Fng}{JI2B1CnfRmePIbNYlckYXw}{x.x.x.x}{x.x.x.x:9300}{dim}{aws_availability_zone=us-west-2c, ml.machine_memory=16040165376, ml.max_open_jobs=20, xpack.installed=true}
   Hot threads at 2019-08-21T03:09:06.548Z, interval=500ms, busiestThreads=3, ignoreIdleThreads=true:

   2.2% (10.9ms out of 500ms) cpu usage by thread 'elasticsearch[c91da54d3211][transport_worker][T#2]'
 8/10 snapshots sharing following 3 elements
   io.netty.util.concurrent.SingleThreadEventExecutor$5.run(SingleThreadEventExecutor.java:906)
   io.netty.util.internal.ThreadExecutorMap$2.run(ThreadExecutorMap.java:74)
   java.base@12.0.1/java.lang.Thread.run(Thread.java:835)

::: {604e4c24815e}{48qPgv5xRBq5enV5WVlzrw}{R4US2aXvSJWS3Jxv_Ky5Ig}{x.x.x.x}{x.x.x.x:9300}{dim}{aws_availability_zone=us-west-2c, ml.machine_memory=16040214528, ml.max_open_jobs=20, xpack.installed=true}
    Hot threads at 2019-08-21T03:09:07.419Z, interval=500ms, busiestThreads=3, ignoreIdleThreads=true:

::: {master_node_1}{GchlCXFpRPi51JfQcP-N6A}{a9w0jMt4S4Wmc54BvROfoQ}{x.x.x.x}{x.x.x.x:9300}{dim}{aws_availability_zone=us-west-2b, xpack.installed=true}
   Hot threads at 2019-08-21T03:08:30.014Z, interval=500ms, busiestThreads=3, ignoreIdleThreads=true:

   16.2% (80.7ms out of 500ms) cpu usage by thread 'elasticsearch[master_node_1][management][T#4]'
 2/10 snapshots sharing following 8 elements
   app//org.elasticsearch.action.support.broadcast.node.TransportBroadcastByNodeAction$BroadcastByNodeTransportRequestHandler.messageReceived(TransportBroadcastByNodeAction.java:391)
   app//org.elasticsearch.transport.RequestHandlerRegistry.processMessageReceived(RequestHandlerRegistry.java:63)
   app//org.elasticsearch.transport.InboundHandler$RequestHandler.doRun(InboundHandler.java:267)
   app//org.elasticsearch.common.util.concurrent.ThreadContext$ContextPreservingAbstractRunnable.doRun(ThreadContext.java:758)
   app//org.elasticsearch.common.util.concurrent.AbstractRunnable.run(AbstractRunnable.java:37)
   java.base@12.0.1/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1128)
   java.base@12.0.1/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:628)
   java.base@12.0.1/java.lang.Thread.run(Thread.java:835)
 8/10 snapshots sharing following 9 elements
   java.base@12.0.1/jdk.internal.misc.Unsafe.park(Native Method)
   java.base@12.0.1/java.util.concurrent.locks.LockSupport.parkNanos(LockSupport.java:235)
   java.base@12.0.1/java.util.concurrent.LinkedTransferQueue.awaitMatch(LinkedTransferQueue.java:740)
   java.base@12.0.1/java.util.concurrent.LinkedTransferQueue.xfer(LinkedTransferQueue.java:684)
   java.base@12.0.1/java.util.concurrent.LinkedTransferQueue.poll(LinkedTransferQueue.java:1374)
   java.base@12.0.1/java.util.concurrent.ThreadPoolExecutor.getTask(ThreadPoolExecutor.java:1053)
   java.base@12.0.1/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1114)
   java.base@12.0.1/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:628)
   java.base@12.0.1/java.lang.Thread.run(Thread.java:835)

::: {1f0f1140e3eb}{DoZG-S6bRQ2TdY_n3INSNg}{bA1Q1-YiRbKO333wW0-iVQ}{x.x.x.x}{x.x.x.x:9300}{dim}{aws_availability_zone=us-west-2c, ml.machine_memory=16040214528, ml.max_open_jobs=20, xpack.installed=true}
   Hot threads at 2019-08-21T03:09:05.670Z, interval=500ms, busiestThreads=3, ignoreIdleThreads=true: