16 Aug
2022
16 Aug
'22
1:23 a.m.
Just to add, it seemed like, previously the stunnel service was connecting to single backend service node via load balancer, even though it was supposed to connect to all backends in round robin mode.
After recent O/S or other update, stunnel service is now connecting to multiple backends in round robin mode (the way it is supposed to do) and is throwing frequent (2+ out of ~10) TIMEOUT connect or closed exceeds errors.