https://rancher.com/ logo
Join Slack
Powered by
This message was deleted.
# harvester
a
This message was deleted.
p
For the longhorn UI specifically, you would have to expose the service 
longhorn-frontend
in the 
longhorn-system
namespace You can change the type of the service from ClusterIP to NodePort, or create an ingress
spec: ingressClassName: nginx rules: - host: URL you want it to be accessible on http: paths: - backend: service: name: longhorn-frontend port: number: 80 path: / pathType: ImplementationSpecific
t
Why is this not a default setup? Because not all the longhorn settings are accessible through harvester as far as i can see. Or should I use the cli for those settings?
p
Which settings in particular? I only had to expose LH for some initial troubleshooting, everything I could need to do can be done through the Harvester/Rancher UI, no?
(oh, except the replica count, which I agree is stupid to lock into LH)
Actually nevermind
Yeah, that's a lot of settings to leave locked away
t
Yeah and the data locality settings I think
p
It's probably assumed that the defaults are good enough for an out of the box experience
Though I had to change replica count because I only have 2 nodes :P
t
Yeah I wanted local nvme speeds for a certain vm and started to find these setting limits as well. I thanks ill look into exposing the longhorn ui
p
It's fairly straightforward, just need terminal access to the harvester node
If you need any help, just ping me
t
Thanks!
w
151 Views
Open in Slack
PreviousNext
Powered by