[
https://issues.apache.org/jira/browse/FLINK-15000?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16993353#comment-16993353
]
fa zheng commented on FLINK-15000:
----------------------------------
Another problem in large parallelism. The request url to get watermark metric
seems to too long to get response. The request url is:
[http://0.0.0.0:8081/jobs/:jobid/vertices/:vertexid/metrics?get=0.currentInputWatermark,......,1199.currentInputWatermark]
Actually, it reminds Unable to load requested file /bad-request when
parallelism is over 156.
> Flink WebUI Metrics is very slow in large parallelism
> -----------------------------------------------------
>
> Key: FLINK-15000
> URL: https://issues.apache.org/jira/browse/FLINK-15000
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / Metrics
> Affects Versions: 1.9.0
> Reporter: fa zheng
> Priority: Major
> Fix For: 1.11.0
>
>
> metrics in web ui are very slow when parallelism is huge. It's hard to add
> metric and choose one metric. I run carTopSpeedWindowingExample with command
> {code:java}
> //代码占位符
> flink run -m yarn-cluster -p 1200 examples/streaming/TopSpeedWindowing.jar
> {code}
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
