This message was deleted.

Do you have SSH access to the node, or have its kubeconfig available for

kubectl

commands?

Yes to both.

I think just the bottom part should have the status

Ok, I think this is it.

Status:

Conditions: Last Update Time: 2023-09-11T145316Z Message: Modified(1) [Cluster fleet-local/local]; alertmanager.monitoring.coreos.com cattle-monitoring-system/rancher-monitoring-alertmanager modified {“spec”{“listenLocal”false,“paused”false,“resources”{“limits”{“cpu”“1000m”}}}}; prometheus.monitoring.coreos.com cattle-monitoring-system/rancher-monitoring-prometheus modified {“spec”{“enableAdminAPI”false,“listenLocal”false,“paused”false,“resources”{“limits”{“cpu”:“1000m”}}}} Status: False Type: Ready Last Update Time: 2023-09-11T145317Z Status: True Type: Processed Display: Ready Clusters: 0/1 State: Modified Max New: 50 Max Unavailable: 1 Max Unavailable Partitions: 0 Observed Generation: 3 Partitions: Count: 1 Max Unavailable: 1 Name: All Summary: Desired Ready: 1 Modified: 1 Non Ready Resources: Bundle State: Modified Modified Status: API Version: monitoring.coreos.com/v1 Kind: Alertmanager Name: rancher-monitoring-alertmanager Namespace: cattle-monitoring-system Patch: {“spec”{“listenLocal”false,“paused”false,“resources”{“limits”{“cpu”“1000m”}}}} API Version: monitoring.coreos.com/v1 Kind: Prometheus Name: rancher-monitoring-prometheus Namespace: cattle-monitoring-system Patch: {“spec”{“enableAdminAPI”false,“listenLocal”false,“paused”false,“resources”{“limits”{“cpu”:“1000m”}}}} Name: fleet-local/local Ready: 0

Hmm, I wonder how these drifted in the first place 🤔 You could try applying its suggested patch, for example to the Prometheus one, and then see if Prometheus is absent from the next

kubectl describe bundle ...

invocation:

kubectl patch prometheus rancher-monitoring-prometheus -n cattle-monitoring-system -p='{"spec":{"enableAdminAPI":false,"listenLocal":false,"paused":false,"resources":{"limits":{"cpu":"1000m"}}}}' --type=merge

Thanks. The prometheus patch helped, but patching the alert manager didn’t seem to make a difference. It applied the patch ok, it just still shows up in the Status message.

Message:           Modified(1) [Cluster fleet-local/local]; <http://alertmanager.monitoring.coreos.com|alertmanager.monitoring.coreos.com> cattle-monitoring-system/rancher-monitoring-alertmanager modified {"spec":{"listenLocal":false,"paused":false,"resources":{"limits":{"cpu":"1000m"}}}}

Yeah, there's a server that Harvester installations ping to see if there are new releases available. I think we haven't updated that server's response to include 1.2.0. Let me confirm with the team to make sure my understanding is correct.

Ah, ok. I was wondering if it was something like that.

Yeah, but it will basically just do the equivalent of this:

kubectl apply -f <https://releases.rancher.com/harvester/v1.2.0/version.yaml>

(in case you wanted to get started on the upgrade before the version server is updated)

Super. Thank you!