Home → General →

Publishing Issue

• Edit

Kris

13 Apr, 2011 01:23 PM

For as long as I can remember (multiple versions), we have had an issue where once something is published, randomly the publish job will finish just fine but it will remain in the queue at 99% for 5 minutes or so before finally being removed. This wasn't a huge deal, I would just see one stuck and delete it knowing everything was published.

After upgrading to 6.8, we still have the issue of things getting stuck at 99% but when I hit delete, it just puts canceled next to the job but doesn't remove it.

First question is if anyone has had this 99% issue and if so, is there a way to fix it? Second question is how do I get this canceled job out of the way?

Thanks,
Kris

1. 1 Posted by Lee Roberson (F... on 13 Apr, 2011 05:29 PM

   

   Kris,

   I've had an open ticket with support about this for quite some time (many many months) but no progress has been made on it.

   I hypothesize that in our case, most of the time it was related to moving a publish job at exactly the "wrong" moment into the top most queue slot, just as publish jobs were transitioning states. But I have also seen it during times when I'm fairly certain no one was doing that.

   • Edit

2. 2 Posted by Kris on 13 Apr, 2011 05:39 PM

   

   Thanks for the response Lee. Maybe by me bringing it up again it will get some movement on your issue as well.

   I did find that stopping and restarting Cascade got rid of the canceled job that was stuck but I would really like to avoid having to do that on my production instance all the time.

   • Edit

3. Support Staff 3 Posted by Tim on 21 Jun, 2011 12:54 PM

   

   Hi guys,

   The behavior you mention above is likely related to this issue. We are working to correct this as soon as possible so be sure to keep an eye out for the release notes of any new patches we make available. As both of you mentioned, this appears to have been a problem for multiple versions perhaps dating back to the 5.x series. We were finally able to reproduce this reliably enough to the point where we could track down the issue.

   Thanks!

   • Edit

4. Tim closed this discussion on 21 Jun, 2011 12:54 PM.