1375670 - rule required signoffs are calculated wrong

Status: RESOLVED FIXED

(Release Engineering Graveyard :: Applications: Balrog (backend), enhancement, P1)

Description

[[e-mail] (:bhearsum)]

We discovered this after initially enabling signoffs for Firefox Beta's. We calculate these in https://github.com/mozilla/balrog/blob/c4f383f7acd0bf65805b33270d8c0a05240bf663/auslib/db.py#L1465 after passing in the current and new version of the rule. However, we don't end up passing the entire current version of the Rule in https://github.com/mozilla/balrog/blob/c4f383f7acd0bf65805b33270d8c0a05240bf663/auslib/db.py#L1720 -- we only pass the product. This ends up with getPotentialRequiredSignoffs thinking that the Rule applies to all channel - so any signoff requirements that match the product will apply to the change.

We're disabling signoffs for now, so this is not blocking, but needs to get fixed ASAP.

Comment 1

[[e-mail] (:bhearsum)]

Attachment: Use full version of current rule when looking up required signoffs for updates

Comment 2

[[github robot]]

Commit pushed to master at https://github.com/mozilla/balrog

https://github.com/mozilla/balrog/commit/eb35f9938d327719edc2fb330edc9476378caf41
bug 1375670: Use full version of current rule when looking up required signoffs for updates (#343). r=rail

Comment 3

[Aki Sasaki (not active)]

I believe this led to fallout for 55.0b4 where I had to remove/manually bump the 55.0b3 rules. If we're planning on leaving the automation as-is for a while, we may need to address this.

Comment 4

[[e-mail] (:bhearsum)]

(In reply to Aki Sasaki [:aki] from comment #3)
> I believe this led to fallout for 55.0b4 where I had to remove/manually bump
> the 55.0b3 rules. If we're planning on leaving the automation as-is for a
> while, we may need to address this.

What exactly happened? I really doubt this bug affected anything after the required signoffs were removed...

Comment 5

[[e-mail] (:bhearsum)]

We tried things again with this patch and everything went well this time!