[
https://issues.apache.org/jira/browse/KYLIN-4857?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17258966#comment-17258966
]
ASF GitHub Bot commented on KYLIN-4857:
---------------------------------------
hit-lacus edited a comment on pull request #1538:
URL: https://github.com/apache/kylin/pull/1538#issuecomment-754652429
Following is my suggestion:
You are using event flow : **SparderMetricsListener** ->
**QuerySparkMetrics** -> **QueryService/QueryMetricsFacade** -> MetricsSystem,
I think it is OK if you send MetricsEvent direct to MetricsSystem :
**SparderMetricsListener** -> **QuerySparkMetrics** -> **MetricsSystem**.
In the previous implementation, when a query is finished, all metrics data
are collected because query and metrics collection are in the same thread. But
in the new implementation, SparkListener and QueryService are not in the same
thread, so the order is not guaranteed . I guess you can call
`MetricsManager.update` to send metrics event in `QuerySparkMetrics` directly .
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
> Refactor system cube for kylin4
> -------------------------------
>
> Key: KYLIN-4857
> URL: https://issues.apache.org/jira/browse/KYLIN-4857
> Project: Kylin
> Issue Type: Improvement
> Components: Metrics
> Affects Versions: v4.0.0-alpha
> Reporter: Yaqian Zhang
> Assignee: Yaqian Zhang
> Priority: Minor
> Fix For: v4.0.0-beta
>
>
> With the change of query engine and storage, the query related metrics
> collected by system cube in kylin3 are no longer applicable to kylin4, so it
> needs to be refactor to adapt to the spark query engine in kylin4.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
