Today I had an issue with updating / creating a new load balancer, the load balancer didn’t finish updating since some of the containers where in “in progress” state.
The containers who where in this state where old and unused containers in which I removed but they kept appearing in the UI.
Only after I restarted the Rancher server these containers started to delete.
How can I make sure it doesn’t happen again ?
Why did the containers go stuck like that and why did they affect the haproxy / network of the whole system?
Were you trying to create a new load balancer or update one? How many containers did you start in your load balancer service? Were the “in progress” state containers the ones in the load balancer service or other services that you were trying to load balancer?
Containers will continue to try and spin up again even if removed from the UI if you have specified a specific scale of the service. Are you saying that even though you deleted containers, those exact containers started up again?
This problem is pretty urgent. It crashed our production few times by now and it seems that we will need to give up Rancher if we do not resolve this. Do you have a payed support? We really believe in your platform but there is a big difference between the theory and how it runs on a real production.
Please get back to me ASAP so you can help my DevOps directly.