I hope this is not to much of old post bump. I just want to say @Levitance Post was amazing and decided to go with LVM route and it was fine till today and wondering if you have experience with this.
So I have a small cluster of 3 machines. I have a feeling the Main machine can’t see the new drive added to the third machine.
This what I did on machine 3:
cfdisk /dev/sdb
pvcreate /dev/sdb1
vgcreate small-sdd-3 /dev/sdb1
Then I went to Datacenter / Storage and Add: LVM and the Volume Group does not show small-ssd-3.
I tried this pvscan on the main server and does not show the other servers drives. I also tried /etc/init.d/pve-cluster restart on both servers without luck.
I have a feeling main cluster reboot would fix the issue but at this time I can’t shutdown / do a full reboot. I mean one the reason to use proxmox and such is to be able create vm’s and ct’s without the need to reboot and I feel like this might be a over looked and silly bug.
Just to add to this I actually did this entire thing fine on the main (1st) machine without issues at all. So for some reason machine 3 just not sending the new info. So now I can’t create a new ct on machine 3 to use the new SSD sadly…
Thank you so much.
Oh look. Someone is having trouble with LVM.
I’ll be eagerly watching to see what the resolution is.
Wait, so you’ve got 3 machines, and you have an example of where local storage was setup on one machine, and shows up on all 3? Because clustered machines generally need a little extra magic to get shared storage working.
Very often you’ll see networked storage like NFS, or if there’s a cluster aware file system in use, iSCSI.
For servers that have local disks, you’ll use distributed replicated block device (DRBD). DRBD requires quite a bit of planning to setup. It needs local disks to exist on all nodes in the cluster, it needs different partitions to be the master, depending on which node in the cluster is normally running the virtual machines in question. It’s like seven different layers of harsh flakes to properly deploy.
Sorry if this is a disjointed post. I just drove 700 miles, and now it’s 1am. Ultimately I think I’ll need to know a little more about the setup. Do you have an existing shared, local storage setup distributed successfully across all 3 nodes? Does your new small-sdd-3 volume group show up on the machine that it was created on? Does it show up on either of the other nodes?
I feel like an idiot here. It was actually really simple fix I don’t know if it’s considered a workaround or not… I just logged in directly to server box 3 and added the new LVM storage through it and worked perfectly… I feel like I’m the biggest idiot in the world lol.