Hi everyone, I’m using Temporal, deployed by a helm chart, I use PostgreSQL as a persistence storage, so I faced an issue that, despite my PostgreSQL instance works fast (I’ve explored the metrics of persistence latency), there is a huge gap between two different activities (and also between workflow start and the first activity), average gap time is 1s, but in different cases it reaches even more. Where can the problem be and how can I fix it? Maybe, there is a mistake in configuration of a helm? But I remain everything as it was (numHistoryShards is 512 and maxConns is 20, which I consider is enough to satisfy at least one workflow). I can’t move my persistence storage to the Temporal instance (it is located at AWS and must be secure and bulletproof), also I can’t use Cassandra, but I don’t think if there is a problem in persistence storage anyway. Thank you! Looking for your answer!
Here is a screenshots of the long gaps:
