Hide Forgot
Created attachment 559141 [details] /var/log/pulp/pulp.log Description of problem: I managed to get the system into a state where all promotions where failing, regardless of whether the content was [un]synced. 16:25:40 ehelms: so the issue I saw in your logs, which looks like we should file a bug on, if during or after a sync, the reindex for elasticsearch fails due to a pulp timeout, the data within katello for repos will be inconsistent Version-Release number of selected component (if applicable): * katello-0.1.211-2.el6.noarch * pulp-0.0.263-1.el6.noarch Steps to Reproduce: 1. Import manifest, enable repositories, and sync some of the enabled repositories 2. Attempt a promotion of the product (which includes some unsynced repositories) 3. Sync remaining unsynced repositories 4. Attempt another promotion Actual results: All promotions failed (see screenshot). Expected results: No failed promotions Additional info:
Created attachment 559142 [details] /var/log/katello/delayed_job.log
I *think* this is no longer relevant, please re-evaluate.
Justin, this is a mighty old one. Given all of the changes that have gone in since (e.g. pulp2.4, dynflow, enginification...etc), have you seen anything like this occurring? In the past, a big contributor to the timeout was resource (e.g.memory).
I believe it is still possible for this behaviour to occur but is much much less likely to with pulp 2.4. I am guessing this will not be reproducible. Will Move to on_qa
Considering this verified, I cannot get it to occur in latest builds. Verified in Satellite-6.0.3-RHEL-6-20140619.0
This was delivered with 6.0.3, which is the Satellite 6 Beta.