Fix tests on postgresql#3740
Conversation
| allow_none=True, | ||
| desc="user_last_seen_monthly_active", | ||
| )) | ||
| defer.returnValue(res) |
There was a problem hiding this comment.
What's the reason for converting all these functions to defer.inlineCallbacks?
There was a problem hiding this comment.
me trying things and inadvertently committing it :) -- reverted
| ) | ||
| self.clock = self.hs.get_clock() | ||
| else: | ||
| self._cache_id_gen = None |
There was a problem hiding this comment.
So, fun fact: this now gets set on the slave replication stores as well as the store on master. This is a bit awkward because we do set self._cache_id_gen on the replication stores in synapse/replication/slave/storage/_base.py to something different.
In short, everything is awful and is scary and I'm scared about moving this here in case the replication stuff fail to properly overwrite this definition
erikjohnston
left a comment
erikjohnston
left a comment
There was a problem hiding this comment.
This looks legit. I've not gone through all the test changes, but they seem sane.
2 participants
No description provided.