brightsky
brightsky copied to clipboard
Investigate deadlocks during source exports
Since e1c4f39 (I think?) we've been getting some deadlock errors from Postgres during update_source()
. But shouldn't our thread lock on DBExporter
prevent this?