You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hbase.apache.org by "Xu Cang (JIRA)" <ji...@apache.org> on 2018/04/13 23:39:00 UTC

[jira] [Updated] (HBASE-20233) [metrics] Ill-formatted numRegions metric in "Hadoop:service=HBase,name=RegionServer,sub=Regions" mbean

     [ https://issues.apache.org/jira/browse/HBASE-20233?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Xu Cang updated HBASE-20233:
----------------------------
    Attachment: HBASE-20233.patch

> [metrics] Ill-formatted numRegions metric in "Hadoop:service=HBase,name=RegionServer,sub=Regions" mbean
> -------------------------------------------------------------------------------------------------------
>
>                 Key: HBASE-20233
>                 URL: https://issues.apache.org/jira/browse/HBASE-20233
>             Project: HBase
>          Issue Type: Bug
>          Components: metrics
>            Reporter: stack
>            Assignee: Xu Cang
>            Priority: Trivial
>              Labels: beginner
>         Attachments: HBASE-20233.patch
>
>
> Noticed this poking at metrics. The MBean "Hadoop:service=HBase,name=RegionServer,sub=Regions" carries per region metrics. They are all formatted like so:
> {code}
> Namespace_default_table_IntegrationTestBigLinkedList_metric_incrementTime_98th_percentile: 0,
> Namespace_default_table_IntegrationTestBigLinkedList_metric_incrementTime_99th_percentile: 0,
> Namespace_default_table_IntegrationTestBigLinkedList_metric_incrementTime_99.9th_percentile: 0,
> Namespace_default_table_IntegrationTestBigLinkedList_metric_appendTime_num_ops: 0,
> Namespace_default_table_IntegrationTestBigLinkedList_metric_appendTime_min: 0,
> Namespace_default_table_IntegrationTestBigLinkedList_metric_appendTime_max: 0,
> Namespace_default_table_IntegrationTestBigLinkedList_metric_appendTime_mean: 0,
> {code}
> In the middle of them all is a metric named...
> {code}
> numRegions: 15,
> {code}
> It has a different format and it is out of scope when compared to the other metrics in this mbean; it belongs better elsewhere, perhaps in the Server bean.
> I noticed it because it breaks the parse done by tcollector scraping our metrics from /jmx
> It was added long ago in HBASE-14166



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)