[jira] [Commented] (CONNECTORS-885) Startup sequence also handles ServiceInterruption badly

["Karl Wright (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/CONNECTORS-885?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13894241#comment-13894241 ] 

Karl Wright commented on CONNECTORS-885:
----------------------------------------

JobManager.getJobsReadyForStartup() is called to locate the jobs in need of processing.  This looks for jobs in the READYFORSTARTUP and READYFORSTARTUPMINIMAL states.  The StartupThread class take care of the actual processing, and when done puts the job into various ACTIVE states via JobManager.noteJobStarted(), or ABORTINGSTARTINGUP via JobManager.errorAbort().  So instead we'll need a JobManager.retryStartup() and the corresponding changes.


> Startup sequence also handles ServiceInterruption badly
> -------------------------------------------------------
>
>                 Key: CONNECTORS-885
>                 URL: https://issues.apache.org/jira/browse/CONNECTORS-885
>             Project: ManifoldCF
>          Issue Type: Bug
>          Components: Framework crawler agent
>    Affects Versions: ManifoldCF 1.6
>            Reporter: Karl Wright
>            Assignee: Karl Wright
>             Fix For: ManifoldCF 1.6
>
>
> The job startup sequence doesn't handle ServiceInterruption conditions fully; it just waits internally and retries.  It should instead make use of the new failtime and failcount job fields to handle them fully.



--
This message was sent by Atlassian JIRA
(v6.1.5#6160)