disk overflow, because failed ingests are not removed

Steps to reproduce

Steps to reproduce:
1. After was fixed, it turned out that the refused ingest because of "WARN (IngestRestService:416) - Workflow {197} is already in operation #10, so we can not ingest" were still kept in the ingest-temp dir. And after the disk overflow because of this the servers halted:
see the logs here http://pastebin.com/gJeS2bxJ

Activity

Show:
Rüdiger Rolf
May 17, 2011, 6:35 AM

I currently try to merge this to our local 1.1 branch, but currently it seems that there have been several changes to the REST endpoint in the meantime (introduction of new constants and changes in the restdocs), so that it's not an easy svn merge.

Josh Holtzman
May 17, 2011, 2:03 PM

Rüdiger, you can use only the changes to IngestRestService.addZippedMediaPackage() and all of the changes to IngestServiceImpl. The rest of the REST stuff doesn't need to be merged.

A
May 17, 2011, 4:22 PM
Edited

As discussed today, after you confirm the changes work in your production environment, please merge these changes into 1.1.x by eod tomorrow

Rüdiger Rolf
May 18, 2011, 2:40 PM

I tried to backport the patch to 1.1.x and added it to our production system... we'll see if this problem will not happen again.

Greg Logan
March 12, 2013, 3:56 PM

I'm going to resolve this since work on 1.2 has long since stopped. If this is really causing you problems please upgrade to 1.4 or newer.

Assignee

Unassigned

Reporter

Rüdiger Rolf

Severity

Data Loss/Corruption

Tags (folksonomy)

None

Components

Fix versions

Affects versions

Priority

Critical
Configure