jackrabbit-oak-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Chetan Mehrotra (JIRA)" <j...@apache.org>
Subject [jira] [Resolved] (OAK-4563) Specify thread pool name which should be used by Async Indexing task
Date Wed, 20 Jul 2016 05:35:20 GMT

     [ https://issues.apache.org/jira/browse/OAK-4563?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Chetan Mehrotra resolved OAK-4563.
----------------------------------
       Resolution: Fixed
    Fix Version/s: 1.5.6

Currently we use following types of periodic jobs

# Async indexing - (Cluster Singleton)
# Document Store - Journal GC (Cluster Singleton)
# Document Store - LastRevRecovery
# Statistic Collection - For timeseries data update in ChangeProcessor,
    SegmentNodeStore GCMonitor

With this change #1, #2 and #3 would use a dedicated pool. While others would use a default
pool

Applied the patch
* trunk - [1753444|http://svn.apache.org/r1753444]

*Usage*
Pool name used is {{oak}}. In Sling this name needs to be added to OSGi config for _Apache
Sling Scheduler_ under _Allowed Thread Pools_ setting

> Specify thread pool name which should be used by Async Indexing task
> --------------------------------------------------------------------
>
>                 Key: OAK-4563
>                 URL: https://issues.apache.org/jira/browse/OAK-4563
>             Project: Jackrabbit Oak
>          Issue Type: Improvement
>          Components: core
>            Reporter: Chetan Mehrotra
>            Assignee: Chetan Mehrotra
>            Priority: Minor
>              Labels: candidate_oak_1_0, candidate_oak_1_2, candidate_oak_1_4
>             Fix For: 1.6, 1.5.6
>
>         Attachments: OAK-4563-v1.patch
>
>
> While running Oak in Sling we rely on Sling Scheduler to ensure that async indexing task
are run on leader (OAK-1246) with specified frequency. 
> Be default Sling Scheduler uses a default pool for managing all tasks. It can happen
that number of task can be quite hight which can then lead to default thread pool getting
exhausted and that causes async indexing to get delayed.
> To ensure that async indexing is not affected by such scenarios we should make use of
a dedicated thread pool. This is now supported by SLING-5831



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

Mime
View raw message