[
https://issues.apache.org/jira/browse/FLINK-6295?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15972841#comment-15972841
]
ASF GitHub Bot commented on FLINK-6295:
---------------------------------------
Github user WangTaoTheTonic commented on the issue:
https://github.com/apache/flink/pull/3709
In my opinion EGHolder is simply a cache which should not be assigned too
complicated task.
If we add the check logic, how long it should be? Will other events
afftects status of tasks? I believe there're more concerns if we added it.
This fix only change internal data structures and decouple with both
JobManager and web frontend.
I am not sure why we are reducing usage of guava, but it sounds not a very
good idea :(
> use LoadingCache instead of WeakHashMap to lower latency
> --------------------------------------------------------
>
> Key: FLINK-6295
> URL: https://issues.apache.org/jira/browse/FLINK-6295
> Project: Flink
> Issue Type: Bug
> Components: Webfrontend
> Reporter: Tao Wang
> Assignee: Tao Wang
>
> Now in ExecutionGraphHolder, which is used in many handlers, we use a
> WeakHashMap to cache ExecutionGraph(s), which is only sensitive to garbage
> collection.
> The latency is too high when JVM do GC rarely, which will make status of jobs
> or its tasks unmatched with the real ones.
> LoadingCache is a common used cache implementation from guava lib, we can use
> its time based eviction to lower latency of status update.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
