[
https://issues.apache.org/jira/browse/FLINK-11912?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16792155#comment-16792155
]
Shuyi Chen edited comment on FLINK-11912 at 3/14/19 12:22 AM:
--------------------------------------------------------------
Hi [~tzulitai], I've attached a proposed tentative change (experimental)
[here|https://github.com/apache/flink/commit/c37394acc01ea5a0c4e2681319ecbfaa63beead3],
could you please take a look and let me know if this is the right approach?
Thanks a lot.
was (Author: suez1224):
Hi [~tzulitai], I've attached a proposed tentative change (experimental)
[here|https://github.com/apache/flink/commit/094135efcadf5c0ddb47eabd66091e20d26d1417],
could you please take a look and let me know if this is the right approach?
Thanks a lot.
> Expose per partition Kafka lag metric in Flink Kafka connector
> --------------------------------------------------------------
>
> Key: FLINK-11912
> URL: https://issues.apache.org/jira/browse/FLINK-11912
> Project: Flink
> Issue Type: New Feature
> Components: Connectors / Kafka
> Affects Versions: 1.6.4, 1.7.2
> Reporter: Shuyi Chen
> Assignee: Shuyi Chen
> Priority: Major
>
> In production, it's important that we expose the Kafka lag by partition
> metric in order for users to diagnose which Kafka partition is lagging.
> However, although the Kafka lag by partition metrics are available in
> KafkaConsumer after 0.10.2, Flink was not able to properly register it
> because the metrics are only available after the consumer start polling data
> from partitions. I would suggest the following fix:
> 1) In KafkaConsumerThread.run(), allocate a manualRegisteredMetricSet.
> 2) in the fetch loop, as KafkaConsumer discovers new partitions, manually add
> MetricName for those partitions that we want to register into
> manualRegisteredMetricSet.
> 3) in the fetch loop, check if manualRegisteredMetricSet is empty. If not,
> try to search for the metrics available in KafkaConsumer, and if found,
> register it and remove the entry from manualRegisteredMetricSet.
> The overhead of the above approach is bounded and only incur when discovering
> new partitions, and registration is done once the KafkaConsumer have the
> metrics exposed.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
