Initial attempt of the subject, created mostly by using ResultsDB as a guide.
Details
- Reviewers
tflink - Maniphest Tasks
- T442: ansible-ize execdb
Untested, I'm not really sure how could I test it, though
Diff Detail
- Lint
Lint Skipped - Unit
Unit Tests Skipped
overall, it looks pretty good to me but I'm wondering if we really want/need execdb to be its own host or if it'd be better to co-habitate the resultsdb VM. thoughts?
Yep, I think it should not be a problem - not sure what would be better - whether resultsdb VM or the VM with Trigger, though. I do not think that it needs its separate VM.
Trigger is on the buildmaster VM, so I think that the resultsdb VM would be a better choice since there are fewer things running on it. Let's just put in on there for now - if it causes problems, we can extract to a separate VM later.
in the spirit of @tflink's comment, I (hopefully) changed the configs, so execdb shares the machine with resultsdb
overall, looks pretty good to me. I think it'd be better to add the execdb role to the existing resultsdb playbook/group instead of trying to control the same VM with multiple playbooks, though.
playbooks/groups/execdb-dev.yml | ||
---|---|---|
2 ↗ | (On Diff #856) | If execdb is residing on the resultsdb machine, it doesn't need a separate playbook - that can cause oddness when things are repeated. I think it'd be better to add the role on to the resultsdb playbook instead of creating another group |