Upgrade salt from 2014.1.11 to 2014.7.5 on all production hosts
Description
Status | Subtype | Assigned | Task | ||
---|---|---|---|---|---|
Declined | ArielGlenn | T102039 Disabling agent forwarding breaks dsh based restarts for Parsoid (required for deployments) | |||
Declined | ArielGlenn | T63882 [Trebuchet] Salt times out on parsoid restarts | |||
Declined | yuvipanda | T87803 Cannot use dsh-based restart of parsoid from tin anymore | |||
Resolved | ArielGlenn | T88971 Upgrade salt to 2014.7 (investigating) | |||
Resolved | ArielGlenn | T98580 upgrade salt on production cluster | |||
Resolved | ArielGlenn | T98581 fix trebuchet-trigger (git deploy) publish.runner arguments |
Event Timeline
Salt has been upgraded on all production hosts running precise, trusty or jessie, with a few exceptions.
Salt on tin (deployment host) is not yet updated, pending a fix to the trebuchet-trigger package, needed for new format of arguments to publish.runner
The following hosts were down and so not upgraded:
dysprosium
labvirt1005
labstore1002
so salt on tin was accidentally upgraded late tonight. I have a package built for trebuchet-trigger that I have installed manually on tin so that deployments aren't broken (see /home/ariel for the deb file if needed). Tomorrow I'll go back and do it properly. Note that testing git deploy of test/testrepo on prod gives no information, as there are no longer any hosts that have the right deployment_target grain.