[
https://issues.apache.org/jira/browse/UIMA-3163?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Lou DeGenaro updated UIMA-3163:
-------------------------------
Fix Version/s: future-DUCC
> DUCC webserver (WS) Run Fails count & Lost
> ------------------------------------------
>
> Key: UIMA-3163
> URL: https://issues.apache.org/jira/browse/UIMA-3163
> Project: UIMA
> Issue Type: Bug
> Components: DUCC
> Reporter: Lou DeGenaro
> Assignee: Lou DeGenaro
> Priority: Minor
> Fix For: future-DUCC
>
>
> 1. Run Fails should be a count of Croaks only
> 2. Lost should not be counted as error, and should be retried
> 3. What failures get counted towards killing job when error threshhold is
> exceeded
> For a running Job, there seem to be 2 problems (with respect to WS
> "guessing"). All seems well once job is completed.
> 4. JP Time Run is too big?
> It seems that run start is incorrectly set to the same value as
> initialization start, when if fact it should be set to the same as
> initialization end!
> degenaro@bluej62:/home/ducc/ducc_runtime/logs$ grep copyTime or.log* | grep
> 5219
> or.log:21 Aug 2013 13:58:26,017 INFO OR.ProcessAccounting - copyTimeInit
> 74525 5219 initialization start:2013.08.21 13:45:24 Wed
> or.log:21 Aug 2013 13:58:26,017 INFO OR.ProcessAccounting - copyTimeInit
> 74525 5219 initialization end:2013.08.21 13:58:14 Wed
> or.log:21 Aug 2013 13:58:26,017 INFO OR.ProcessAccounting - copyTimeRun
> 74525 5219 run start:2013.08.21 13:45:24 Wed
> or.log:21 Aug 2013 14:11:25,930 INFO OR.ProcessAccounting - copyTimeRun
> 74525 5219 run start:2013.08.21 13:45:24 Wed
> or.log:21 Aug 2013 14:11:25,930 INFO OR.ProcessAccounting - copyTimeRun
> 74525 5219 run end:2013.08.21 14:11:05 Wed
> or.log.1:21 Aug 2013 13:45:25,934 INFO OR.ProcessAccounting - copyTimeInit
> 74525 5219 initialization start:2013.08.21 13:45:24 Wed
> -----
> For JP in progress, insure that TimeWindowRun start/end values are
> initialized to TimeWindowInit end value.
> Code delivered, part 4 resolved.
> 5. JP %CPU is too small (0?)
> See Jira 3164, part 5 resloved.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
