https://rancher.com/ logo
Join Slack
Powered by
This message was deleted.
# general
a
This message was deleted.
c
go into global settings, edit the metadata config, and reduce the refresh interval
there is no direct way to trigger an on-demand refresh, but changing that and then changing it back should do it
k
So in my case the interval is 1440 minutes (basically every day). Should Rancher download the updated versions then every day? As I don't think that's happening (looking at the repo the 1.33.10 is added 5 days ago)
c
it should happen on that interval, yeah. you can check the rancher pod logs for errors, and ensure that all your rancher instances are pointing at the correct data.json
k
Thanks will take a look. Unfortunately my logs are a bit spammed due to another issue but will figure it out.
Just gave it a try, unfortunatly no luck; 
2026/04/15 19:18:22 [INFO] Skipping handler for clusterrepo rancher-charts. NumberOfRetries is 0, MaxRetry is 3, ClusterRepo Generation is 5, ObservedGeneration is 5, LastUpdated plus interval is 2026-04-15 20:05:28 +0000 UTC, now is 2026-04-15 19:18:22.044025827 +0000 UTC
c
that is not KDM, that is the helm charts repo (under apps)
k
Somehow the reload is not working then.. Searching our logs from the last 24h doesn't state any log 
Loaded configuration from <https://releases.rancher.com/kontainer-driver-metadata/release-v2.13/data.json>
. While this is logged on the instance that shows the 1.33.10.
c
are you able to curl that URL if you exec into the rancher pod?
you can always just try restarting the rancher pod to see if that kicks it
k
curl gives a warning about an insecure cert, but both instances are reporting that error. using the --insecure downloads the data.json file without any issues.
Ill try tomorrow a restart of the pod to see if that changes something. Will update this thread with my conclusion.
Thank you for your help!
Restarting the pod forces indeed a refresh of the metadata.
74 Views
Open in Slack
PreviousNext
Powered by