This message was deleted.

Sometimes stable and latest match, but the 1.4.0 isn't stable yet.

Ok - thanks for making clear! Will keep an eye out.

from the website: https://www.suse.com/suse-harvester/support-matrix/all-supported-versions/harvester-v1-3-x/#:~:text=All%20Harvester%20releases%20will%20go,2%20months%20of%20upgrade%20period%20).

All Harvester releases will go through three stages: In Development, Latest, and then finally Stable. Please ensure you do not upgrade your production cluster to anything other than a listed Stable version. Use Latest versions for test purposes only.

OK - if thats the case - why doesn't it show in the list of upgrades I wonder... at least for me the list is empty...

what version are you on?

yep

Must be you because it's there for me:

this is what i get

click the empty box?

nothing there

If it doesn't pop up clear your cookies and cache

nothing

that's a different browser or incognito window or something?

just trying incognito

The only other thing I can think of is this in the settings:

indeed I think thats the same

See what happens when you go into the UI directly vs via Rancher

Will try that in a mo.. got a broken cluster to sort out first tho... (we had a power cut over xmas and lost power for 2days - UPS wan't going to cover that) VM's all came up ok but K8's cluster was very broken. Hoping this isn't all related of course...

Did all the VMs get the same IP? If it’s DHCP and the VMs have the wrong/different IP, that can happen.

Yes - that we think was the issue and we’ve since reserved the IPs for cluster nodes post assignment. Possibly we need to use a seperate subnet that’s managed within rancher - might test that on a rainy day. The router manages this atm and it’s quite easy to map IPs but it’s a pain as it’s not really proper IAAS is it? Anyway also need to check going to harvester direct for the upgrade availability- which this thread was really about but presently completing evaluating backup and restoring using velero for the cluster - which is going well so won’t be long before I can get back to the upgrade with confidence :)

The issue is all in your IP reservation and k8s needing the same IP.

Urm - yes, this thread however was todo with missing upgrade option to 1.4.0 and assumption it might not be stable (hence why I couldn't see it), the cluster issue mentioned was just about some background, on the IP reservation doing this externally does fly in the face of principles of IAAS as mentioned, but I really think that discussion could be on a fresh thread another day.... I hope to be ready to check if the options show if I login direct to harvester rather than via rancher - which I suspect will fix the issue of not seeing the update (though one maybe forgiven for thinking this would work this way...)