accumulo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Christopher Tubbs (JIRA)" <>
Subject [jira] [Commented] (ACCUMULO-4458) lock contention around configuration settings impacts tablet server performance
Date Thu, 15 Sep 2016 20:04:20 GMT


Christopher Tubbs commented on ACCUMULO-4458:

bq. Is this still unclear from the earlier discussion?

Yes, because of what [~busbey] said:

bq. If we want to do that, it's simpler to just not keep around / use the underlying Hadoop
Configuration object and default to the parent config for anything that's not in the cached
hashmap. I'll push this and then make a follow on to give that a try.

It's not clear that was attempted yet... but I think it's the right way to go.

bq. I believe what he has presently is correct.

You're right. I was confused about the semantics of inserting null values. If we adopt the
strategy I quoted from Sean, then that whole part goes away, of course.

> lock contention around configuration settings impacts tablet server performance
> -------------------------------------------------------------------------------
>                 Key: ACCUMULO-4458
>                 URL:
>             Project: Accumulo
>          Issue Type: Bug
>          Components: tserver
>    Affects Versions: 1.7.1, 1.8.0
>            Reporter: Sean Busbey
>            Assignee: Sean Busbey
>            Priority: Critical
>             Fix For: 1.7.3, 1.8.1, 2.0.0
>         Attachments: ACCUMULO-4458-1.7.v1.patch
>          Time Spent: 50m
>  Remaining Estimate: 0h
> While investigating a pretty severe performance regression comparing YCSB against 1.6
and 1.7, I found a fair bit of lock contention around getting configuration values. This was
improved by ACCUMULO-4388, but various threads eventually all started contending on the configuration
values expected from the site xml files.

This message was sent by Atlassian JIRA

View raw message