[jira] [Updated] (ACCUMULO-2997) Race condition in resizing of TServer client pool

["Corey J. Nolet (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/ACCUMULO-2997?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Corey J. Nolet updated ACCUMULO-2997:
-------------------------------------
    Fix Version/s:     (was: 1.6.2)
                   1.6.3

> Race condition in resizing of TServer client pool
> -------------------------------------------------
>
>                 Key: ACCUMULO-2997
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-2997
>             Project: Accumulo
>          Issue Type: Bug
>          Components: tserver
>    Affects Versions: 1.5.1, 1.6.0
>            Reporter: Josh Elser
>             Fix For: 1.5.3, 1.7.0, 1.6.3
>
>          Time Spent: 0.5h
>  Remaining Estimate: 0h
>
> From comments:
> {noformat}
> there is a race condition here... the active count could be higher by the time
> we decrease the core pool size... so the active count could end up higher than
> the core pool size, in which case everything will be queued... the increase
> case should handle this and prevent deadlock
> {noformat}



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)