Need help moving VMs to shared storage. HA acting weird.

C

CaliMiller

New Member
Mar 23, 2025
13
2
3
Indiana, US
calimiller.me
Hello,

I am rather new to Proxmox VE, and still learning. Nothing is mission-critical, I'm just tinkering with a home lab as a hobby.

I bought a mini PC about 6 months ago and installed Proxmox. After some issues, and some trial and error I got it running smoothly. I later setup a network share on my NAS to act as storage for Proxmox.

Yesterday, I added another node, creating a cluster. On the second node (radserve2), I installed another instance of Pi-Hole (container 2100, pihole3), so there is redundancy.

Today, I added a third node (radserve3) to the cluster, and decided to enable High Availability. I must have done something incorrectly, because container 2100, migrated to the first node RadServe, and I can't get it to migrate back to radserve2, even though radserve2 shows as being online. Container 2100 is also frozen, and won't start when I click to start it.

2025-03-23-Screenshot 0001.png

Also, several of the VMs are stored on the first node, and I'd like to move them to the NAS storage which is shared by the cluster, but when I attempt to do it, it doesn't give me the option to move to the NAS. The instances installed on the node storage were created with Proxmox Helper Scripts.

2025-03-23-Screenshot 0002.png

Also, is there a way to change the node name from RadServe to radserve1, so it doesn't share the name of the cluster?

Thank you for your help.
-Cali
 
Last edited:
Hi @Lukas Moravek

Thank you for your reply. I restarted all of my nodes, and now none of the containers will start. I click the start button, and it records it in the Tasks and the Cluster Log, but nothing happens. No error messages. The Cluster Log shows the task starting and immediately ending. Is there a way to disable high availability on a container or VM?

2025-03-23-Screenshot 0003.png

2025-03-23-Screenshot 0004.png
 
SteveITS said:
Your storage Radnas is all ? icons. HA state is freeze. Is storage working?
Click to expand...

I seem to have created a paradox. The Pi-Holes act as my LocalDNS. But since neither Pi-Hole will open (and Pi-Hole 1 on a dedicated Raspberry Pi is currently non-functional), they can't resolve the domain I set for RadNAS. I should have used the IP address.

1742783731821.png

I've changed that to the IP address, and the question marks have disappeared. However, I still can't seem to start any of the containers or VMs.
 
Last edited:
I looked in another thread and found this way to remove VMs from HA. Once I did this, the VMs and containers will start. I was also able to migrate container 2100 back to radserve2.

t.lamprecht said:
Hi,

you mean remove all services from being HA managed in a single step?

The web UI/API and CLI only allows for removing a service at once, but if you want to remove all you can simply remove the whole config where it's defined which service should be managed by HA, e.g.:
Bash: 
rm /etc/pve/ha/resources.cfg

Note that it might need a few minutes until that went through to all notes, and the run-status of the VMs & CTs themselves should not be affected by this.
Click to expand...

So I guess the question is, what caused everything to not work, so I don't mess it up again in the future?
 
You must log in or register to reply here.
Top Bottom