Gc-young error in ES

sunilmchaudhari · March 21, 2016, 1:08pm

Hi,
can anybody tell me the meaning of this error?

[2016-03-21 14:43:27,809][INFO ][monitor.jvm ] [prod-data-node-2] [gc][young][261567][1099] duration [770ms], collections [1]/[1.1s], total [770ms]/[6.6m], memory [2.1gb]->[2.3gb]/[9.9gb], all_pools {[young] [20.2mb]->[4.5mb]/[266.2mb]}{[survivor] [33.2mb]->[33.2mb]/[33.2mb]}{[old] [2gb]->[2.3gb]/[9.6gb]}

[2016-03-21 14:39:15,749][INFO ][monitor.jvm ] [prod-data-node-1] [gc][young][110003][949] duration [938ms], collections [1]/[1.5s], total s]/[1m], memory [6.1gb]->[5.8gb]/[9.9gb], all_pools {[young] [181.7mb]->[5.1mb]/[266.2mb]}{[survivor] [33.2mb]->[33.2mb]/[33.2mb]}{[old] [5.9gb]->[5.8gb6gb]}

br,
Sunil

danielmitterdorfer · March 21, 2016, 1:17pm

Hi Sunil,

this is no error at all. It just gives you some metrics of young generation garbage collections that have occurred.

Daniel

sunilmchaudhari · March 21, 2016, 1:28pm

Hi,
But I want it to understand, what does it mean?
Can you please explore what it says?

It suddenly cam when one system started sending logs to centralized logging and those have huge log size accumulated sometime back.

danielmitterdorfer · March 21, 2016, 1:35pm

Hi,

these two lines produce this log statement:

github.com

elastic/elasticsearch/blob/25531b7299ff19b56ed939478332b75a4c574dd9/core/src/main/java/org/elasticsearch/monitor/jvm/JvmGcMonitorService.java#L188-L189


logger.info("[gc][{}][{}][{}] duration [{}], collections [{}]/[{}], total [{}]/[{}], memory [{}]->[{}]/[{}], all_pools {}",
        gc.getName(), seq, gc.getCollectionCount(), TimeValue.timeValueMillis(collectionTime), collections, TimeValue.timeValueMillis(currentJvmStats.getTimestamp() - lastJvmStats.getTimestamp()), TimeValue.timeValueMillis(collectionTime), gc.getCollectionTime(), lastJvmStats.getMem().getHeapUsed(), currentJvmStats.getMem().getHeapUsed(), JvmInfo.jvmInfo().getMem().getHeapMax(), buildPools(lastJvmStats, currentJvmStats));

Daniel