mirror of
https://github.com/neondatabase/neon.git
synced 2026-01-08 05:52:55 +00:00
## Problem There is a rare race between controller graceful deployment and shard splitting where we may incorrectly both abort _and_ complete the split (on different pods), and thereby leave no shards at all in the database. Related: #11254 ## Summary of changes - In complete_shard_split, refuse to delete anything if child shards are not found