This is just an update to say nothing new to report. I was on vacation Friday to Wednesday of this week. Back working the issue full-time.
You might want to consider upgrading to an 8 core machine while we continue to work the problem. It would reduce the chances of error. It is not a cure, just a bandaid.
Thank you for the router / gateway info. It supports my colleague's thoughts.
... that said, I have a test that just now caught a heap management error that might be part of your experience. I can only hope. More on this Monday after I review.
Yes, I finally have a repeat of your crash conditions.
We have a fix in test now. Might have a testing release for you Wednesday. It will be a full product release, not just libStarrocks.so. There were changes needed in the Java layer to better handle rollbacks that occur after a broken connection, i.e. after a router decides a connection has been idle too long.
Sounds like great news! Looking forward to try it out!
I have uploaded a recent nightly snapshot to a shared location. If you send me your email in a DM, I'll give you access.
Thanks, I've sent you a DM with my email!