Optimize capistrano post-deploy tasks #1341
Conversation
**Why**: chmod necessary post-deploy but we can cut down on already-fixed files. Restart passenger post-chmod so that web nodes are synced.
| @@ -86,10 +86,17 @@ | |||
| desc 'Modify permissions on /srv/idp' | |||
| task :mod_perms do | |||
| on roles(:web), in: :parallel do | |||
| execute :sudo, :chown, '-R', 'ubuntu:nogroup', deploy_to | |||
| execute :sudo, :chown, '-R', 'ubuntu:nogroup', deploy_to + '/current', '/srv/idp/shared' | |||
There was a problem hiding this comment.
In my experience the chown has occasionally timed out (too many files?) so adding more directories and files to this seems like it would make things worse? And potentially block things after it like passenger:restart ?
There was a problem hiding this comment.
This reduces the effect of the chmod command to only the current directory.
There was a problem hiding this comment.
(and specifically, the shared) instead of all of the /srv/idp
There was a problem hiding this comment.
this change makes fewer files because it is acting on /srv/idp/current /srv/idp/shared rather than all of /srv/idp.
There was a problem hiding this comment.
ACTUALLY PETER, I'm going to cut you off right there 🎙 ....
Peter doesn't always test his PRs, but when he does, he does it while promoting code to production.
There was a problem hiding this comment.
heh. QA'd in qa env.
**Why**: chmod necessary post-deploy but we can cut down on already-fixed files. Restart passenger post-chmod so that web nodes are synced.
Why: chmod necessary post-deploy but we can cut down on
already-fixed files. Restart passenger post-chmod so that web nodes
are synced.