You are viewing a plain text version of this content. The canonical link for it is here.

Posted to hdfs-issues@hadoop.apache.org by "Farhan Khan (Jira)" <ji...@apache.org> on 2020/05/05 22:53:00 UTC

[jira] [Created] (HDFS-15336) GC metrics are not working with Telegraf's prometheus

Farhan Khan created HDFS-15336:
----------------------------------

             Summary: GC metrics are not working with Telegraf's prometheus
                 Key: HDFS-15336
                 URL: https://issues.apache.org/jira/browse/HDFS-15336
             Project: Hadoop HDFS
          Issue Type: Bug
    Affects Versions: 3.2.1
            Reporter: Farhan Khan


The GC related metric particularly GCTimeMillis is reported as counter via StatsD. Due to this when we use Telegraf's Prometheus to observe the metric it shows unexpected behavior. Telegraf's default behavior is to increment the counter.

https://github.com/apache/hadoop/blob/a0da1ec01051108b77f86799dd5e97563b2a3962/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/metrics2/source/JvmMetrics.java#L128

To see the discrepancy in the metrics you could compare the logs reported by JVM with GClog enabled and the metrics at Prometheus interface.

To fix this either it should be changed to gauge or it should support Prometheus format.

As far as I know, HDFS does not support Prometheus yet.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-help@hadoop.apache.org