[
https://issues.apache.org/jira/browse/FLINK-11344?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16744769#comment-16744769
]
BoWang commented on FLINK-11344:
--------------------------------
[~hailong wang] Thanks for the comment.
The execution attempt list will indeed becomes a bit longer if failures occur,
but I think keeps restarting is rare, so a bit longer list may be tolerable.
Displaying failed attempt could give us a glimpse of the running information
without query the log, including running time, on which TaskManager it runs
etc. According these simple informations, we could make some actions, e.g., by
*orderBy host* on dash board, we can also find the bad host on which many
attempts fail to make corresponding operation actions, e.g., add the machine to
blacklist.
Thus, I think displaying all the execution attempts is more good than harm.
> Display All Execution Attempt Information on Flink Web Dashboard
> ----------------------------------------------------------------
>
> Key: FLINK-11344
> URL: https://issues.apache.org/jira/browse/FLINK-11344
> Project: Flink
> Issue Type: Improvement
> Components: Webfrontend
> Reporter: BoWang
> Assignee: BoWang
> Priority: Minor
> Labels: pull-request-available
> Time Spent: 10m
> Remaining Estimate: 0h
>
> Currently, only one Execution Attempt of each sub-task is shown in web
> dashboard, thus, only the succeed Attempt is shown when failover occurs. This
> may be inconvenient to rapidly locate the failure reasons of failed Attempts
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
