1068799 - Docs: Outline the responsibilities of admin node vs worker & various services on each

Status: RESOLVED WONTFIX

(Tree Management :: Treeherder, defect, P4)

Description

[Ed Morley [:emorley]]

It would be great if we could document (either in RTD or else on a wiki page somewhere) the responsibilities of the various machines used in production (and also dev/staging) and what services are run on each + which need restarting for certain issues etc.

I was trying to deploy bug 1035222 today & whilst the script in bug 1067262 would be the best fix (since it would all be done for me), it would help to know the above for situations like this, or future debugging.

Comment 1

[Ed Morley [:emorley]]

(In reply to Ed Morley [:edmorley] from comment #0)
> + which need restarting for certain issues etc.

This is partly covered by:
https://treeherder-service.readthedocs.org/en/latest/common_tasks.html

Comment 2

[Ed Morley [:emorley]]

Wontfix given heroku makes this clearer (named worker types in Procfile, with commands run for each).