This message was deleted.
# rke2a
adamant-kite-43734
02/24/2023, 10:41 PMThis message was deleted.
c
creamy-pencil-82913
02/24/2023, 10:42 PMno downtime to what
creamy-pencil-82913
02/24/2023, 10:43 PMworkload pods keep running while the service is restarted, but the kubelet/apiserver/etc will temporarily go away as they are upgraded.
creamy-pencil-82913
02/24/2023, 10:43 PMIf you follow the docs and upgrade only a single server at a time you should be fine
h
hundreds-airport-66196
02/24/2023, 10:44 PMah, that could be the key. thanks
c
creamy-pencil-82913
02/24/2023, 10:45 PMIs there anything specific you are running into problems with?
h
hundreds-airport-66196
02/24/2023, 10:50 PMno problem yet. Our current upgrade is being run from ansible playbook and doing this simultaneously is causing downtime. A new feature is ask for a no downtime during the upgrade
c
creamy-pencil-82913
02/24/2023, 10:51 PMyeah, you definitely don’t want to just have ansible just do all the nodes at once. Best practice is to do servers one at a time, waiting for everything to stabilize before moving on to the next. Then you can do the agents with as much parallelism as you want.
h
hundreds-airport-66196
02/24/2023, 10:52 PMthanks @creamy-pencil-82913
4 Views