zookeeper-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From karanmehta93 <...@git.apache.org>
Subject [GitHub] zookeeper issue #307: ZOOKEEPER-2770 ZooKeeper slow operation log
Date Thu, 26 Oct 2017 20:04:30 GMT
Github user karanmehta93 commented on the issue:

    Due to some reason, Jenkins didn't build it automatically. I triggered a manual build
with my latest commit ID SHA1 and the results are at https://builds.apache.org/job/PreCommit-ZOOKEEPER-github-pr-build/1123/
    All the Findbugs warnings have been resolved, and the single failing test seems to be
unrelated to my change (or possibly a flapper since it is passing on my local machine).
    This commit is a squashed commit based on the earlier inputs provided by everyone. Thank
    As of now, the default value for `requestWarnThresholdMs` is 10 seconds, which is too
high. Users can set it according to the average latency values they typically see from the
serverstats. The patch also limits logging rate, so if the value is set too low, it will print
out a message every 60 seconds showing the number of high latency requests in past minute.
The motivation of this JIRA is specified in description at  https://issues.apache.org/jira/browse/ZOOKEEPER-2770.
    If you can review this PR, it would be great.
    @hanm @tdunning @eribeiro @skamille 
    If we can get in this JIRA, the future plan is to work on @tdunning t-digest and use it
to determine high latency requests specifically based on its type. Thanks!


View raw message