Donald Stufft
Donald Stufft
Is this issue useful? If so I'm fine keeping it open, but it feels like the kind of meta issue that never actually gets closed because there's no actual end...
I can emit the event in the admin view to fix that oversight!
I've updated this pull request to use a `before_insert` hook and verified that the query taking the advisory lock is still being emitted prior to an `INSERT` of a Journal....
I've also reverted most of the places where we stopped emitting `JournalEntry` items, since the journal is our only place where we durably store these events currently.
> I'm still curious how this will operate at scale, and how we are likely to determine if we're getting lock contention. Is that something you've considered already? If it...
Thanks for picking this up by the way! Really jazzed to see it getting worked on.
I'm trying to think how this could have happened, and I think I have a theory for it. Multiple transactions occurring in isolation, ``READ COMMITTED`` 1. Transaction A inserts record...
I think that https://github.com/pypi/warehouse/pull/13936 might fix this.
With https://github.com/pypi/warehouse/pull/13936 merged now (along with a few follow up PRs to fix some deadlocks), I *think* that the primary cause of this has been fixed now. The tl;dr is...
The other thing we could do, is forcibly normalize ourselves, though we hadn't done that in the past and I know that would break at least twine's checks if a...