[
http://jira.nuxeo.org/browse/NXP-2661?page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#action_24145
]
Thomas Roger logged work on NXP-2661:
-------------------------------------
Author: Thomas Roger
Created on: 16/09/08 18:26
Start Date: 16/09/08 18:26
Worklog Time Spent: 1 day
Issue Time Tracking
-------------------
Time Spent: 1 day
> Make indexing Thread Pool more reliable
> ---------------------------------------
>
> Key: NXP-2661
> URL: http://jira.nuxeo.org/browse/NXP-2661
> Project: Nuxeo Enterprise Platform
> Issue Type: Bug
> Affects Versions: 5.1.4, 5.1.5
> Reporter: Thierry Delprat
> Assignee: Thomas Roger
> Priority: Critical
> Fix For: 5.1.6
>
> Original Estimate: 1 day
> Time Spent: 1 day
> Remaining Estimate: 1 day
>
> Current Indexing Thread pool implementation has al least 2 problems :
> - Thread pool manage both Recursion tasks and indexing tasks
> (this can create deadlocks in case of multiple reindexing)
> - We have no direct control of the task queue and tasks store the
> documentModel
> (this can lead to indexing multiple times the same document, in the wrong
> order)
> This must be refactored :
> - Tasks needs to be as light as possible (only Repo/DocRef)
> - use an Unbounded Priority Lists for indexing tasks queue
> ( make it remove duplicated tasks)
> - Separate Indexing Thread from recusion threads in 2 separated thread Pools
> ( reuse the ReindexAll ThreadPool)
> -
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://jira.nuxeo.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
_______________________________________________
ECM-tickets mailing list
[email protected]
http://lists.nuxeo.com/mailman/listinfo/ecm-tickets
