[
https://issues.apache.org/jira/browse/NIFI-7379?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17094874#comment-17094874
]
ASF subversion and git services commented on NIFI-7379:
-------------------------------------------------------
Commit 1259bd5bd17c9ebecf918b23f31ca83678c7a3b8 in nifi's branch
refs/heads/master from Matt Burgess
[ https://gitbox.apache.org/repos/asf?p=nifi.git;h=1259bd5 ]
NIFI-7379: Support multiple instances of Prometheus registries/metrics (#4229)
* NIFI-7379: Support multiple instances of Prometheus registries/metrics
* NIFI-7379: Refactored Prometheus objects to support multiple instances
> Prometheus components should not use the same registries or metric objects
> --------------------------------------------------------------------------
>
> Key: NIFI-7379
> URL: https://issues.apache.org/jira/browse/NIFI-7379
> Project: Apache NiFi
> Issue Type: Bug
> Reporter: Matt Burgess
> Assignee: Matt Burgess
> Priority: Major
> Time Spent: 40m
> Remaining Estimate: 0h
>
> Currently all Prometheus components in NiFi (the REST endpoint, the reporting
> task, and the record sink) use the same set of metric objects and collection
> registries. This can cause undesired behavior, such as causing label
> conflicts (for different Instance Identifier values for example), undesired
> metrics to be present (if QueryNiFiReportingTask adds metrics,
> PrometheusReportingTask will expose them too), injection of bad data points
> (if you have a bad query that overwrites an existing metric), etc.
> Each component should have its own copy of the collection registries and
> metric objects so as not to interfere with those of other Prometheus
> components.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
