This message was deleted.
# random
a
This message was deleted.
r
Note that 1.24 has PSPs & Rancher uses them but those went away in 1.25 so there's a difference you need to put in config for that. Also is 1.26 supported by Rancher 2.7.2? I've heard people talk about 1.25 but I thought 1.26 wasn't supported quite yet.
w
Interesting I though that limitation was only for the cluster where Rancher is running not for the ones manage. Maybe I miss understood.
r
There's a different spot in the compatibility matrix for each, as I recall.
w
Yes, my bad I miss understood the matrix 😞
r
So if you drop downstream to 1.25 and make sure you have the flag for using PSPs on downstream set to false (which I don't remember) then that may fix you up.
βœ… 1
w
Ty πŸ™ I'm testing that now
So I was able to downgrade the downstream cluster to 1.24 just to make sure they are in the version, and the error is the same 😞
r
How confident are you in the downgrade? Did you completely remove RKE2 and create a new cluster and migrate workloads or something else?
I've not heard a lot of confidence about doing something other than removing & recreating for downgrading versions of Kubernetes.
w
I wiped the cluster entirely so is a brand-new cluster
r
Were you able to successfully delete the old cluster from Rancher first? I've heard of times where Rancher remembers a node even after the node being rebuilt and that getting in the way.
w
I followed these steps: 1. Removed the downstream cluster from Rancher. βœ… 2. Wiped the downstream cluster and re-install RKE2 1.24 βœ… 3. Imported the new downstream using the new yaml from the Rancher βœ…
AgentDeployed	False	15 mins ago	[Error] Error while applying agent YAML, it will be retried automatically: exit status 1, Error from server (NotFound): the server could not find the requested resource
r
I'm out of ideas, hopefully someone else pops in.
πŸ™ 1
You might try re-posting in general to get a response.
(just include a link to this one for others' info)
βœ… 1