ppawel commented on PR #32962: URL: https://github.com/apache/beam/pull/32962#issuecomment-2505468316
@bzablocki We have this PR running in our test environment and so far looks suspiciously good ;) What we tested so far is: 1. Consuming a large number of spooled messages (they spooled since the previous pipeline was shut down until the new one started with the fresh PR version) - that worked fine, consumed almost immediately. 2. Peak in messages - with JmsIO we have a problem that processing and acking those messages during peak evening hours pretty much always gets behind even though there are enough workers to handle it.. so we end up with a large spool. With this current SolaceIO version the spool is basically empty which is our main goal with this particular pipeline. 3. "Business as usual" during the day - this also looks good, pipeline doesn't get behind and messages are not left in the queue unacked. So overall it looks good, we will definitely keep testing this version, especially points (1) and (2) we need to make sure those are stable before we replace JmsIO. The only thing I have so far is these logs right at the startup yesterday:  I'm not sure if it's by design but just seems strange that there are messages in the queue (see above point 1) when this new pipeline starts but for some reason the connections to Solace need to be reestablished after 1 minute of idle time for it to start consuming. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
