You can try clearing the job queue as a last resort and do not try this in production.
–––––––––––––––––––––––
| PROBLEM DESCRIPTION |
–––––––––––––––––––––––
We have a job in the publishing queue in our development environment that
continues to cause the endpoint to be suspended. Is it possible to just clear
out the entire queue?
––––––––––––––
| RESOLUTION |
––––––––––––––
(See also KB 9781 – http://support.vignette.com/VOLSS/KB/View/1,,9781,00.html)
These SQL statements will clear all the jobs. It is
strongly advised to use this only on non–production systems and to make a
backup of the tables that are affected prior to running the scripts.
update vgnJob
set state = 'DpCompleted'
where state in ( 'WfCompleted', 'WfCancelled', 'inprocess', 'invalid',
'initialized' );
delete from vgnJob
where state in ( 'DpSucceeded', 'DpFailed', 'DpCompleted' );
delete from vgnJobItemSeed
where jobId not in (select id from vgnJob);
delete from vgnJobItemAutoStage
where jobId not in (select id from vgnJob);
delete from vgnJobItemOrder
where jobId not in (select id from vgnJob);
delete from vgnJobItemStage
where jobId not in (select id from vgnJob);
delete from vgnJobStageVisited
where jobId not in (select id from vgnJob);
delete from vgnJobEngineFailure
where epId not in (select id from vgnEpInfo);
delete from vgnJobEngineFailure
where jobId not in (select id from vgnJob);
––––––––––––––––––––––––––––––––––––––––––––––––––––
–– Now that we've removed auto–prune jobs, there may
–– be more subjob info to get rid of.
––––––––––––––––––––––––––––––––––––––––––––––––––––
delete from vgnSubJob
where jobId not in (select id from vgnJob);
delete from vgnSubJobHistory
where jobId not in (select id from vgnJob);
delete from vgnSubJobProgress
where jobId not in (select id from vgnJob);
