[
https://issues.apache.org/jira/browse/UIMA-3451?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13839377#comment-13839377
]
Jim Challenger commented on UIMA-3451:
--------------------------------------
I can't duplicate this because in fact the handleModifiedServices method does
not seem to be bypassed. It's not clear this was actually a problem (or *the*)
problem when it was reported.
> DUCC SM Common processing for instance death
> --------------------------------------------
>
> Key: UIMA-3451
> URL: https://issues.apache.org/jira/browse/UIMA-3451
> Project: UIMA
> Issue Type: Bug
> Components: DUCC
> Affects Versions: 1.0-Ducc
> Reporter: Jim Challenger
> Assignee: Jim Challenger
> Fix For: 1.0-Ducc
>
>
> SM should use common processing when OR reports instance death for both
> modified, and for defunct service instances.
> To replicate problem: create an instance that dies in resource allocation so
> it never advances to Initializing. This sends it directly to
> handleDeletedServices in ServiceHandler which does not currently manage
> restart / redrive policies.
> The restart/redrive code in handleModifiedServices should be made common and
> called from both 'handle' methods.
--
This message was sent by Atlassian JIRA
(v6.1#6144)
