Closed gandelman-a closed 7 years ago
This looks like it was caused by stale jobs registered in gearman for nodes that no longer exist. We bounced the zuul/gearman service and rebuilt nodepool slaves and it appears to be working again. Will reopen this if we run into it again.
Jobs are failing quickly with a POST_FAILURE error.
Looking through logs the zuul-launcher appears to be having problems reaching the slaves to do its ansible runs: http://paste.openstack.org/show/605543/