Mounting Cephfs in VM keeps timing out (kernel in fstab)- desperate for help...

brodgers

New Member
May 3, 2023
1
0
1
As the title would indicate, I am trying to mount my ceph filesystem (quincy) on a VM within the cluster (VE 7.4-3). The ceph filesystem, data pool, metadata pool are all up and running, my monitor nodes are up and happy and ceph health is good.

So far, I have - on the nodes:
- created a keyring in /etc/pve/priv for ceph.client.vmuser.keyring
- imported the new keyring
- exported their key to a vmuser.secret file
- scp'd that file to the VM client /etc/ceph
- generated a minimal config
- scp'd the minimal config to the VM client /etc/ceph

On the (Debian 11) client I have:
- installed cep-common
- mkdir /etc/ceph
- created a mount point at /mnt/cephfs
- edited my fstab to be:
Code:
10.66.10.1:6789,10.66.10.4:6789,10.66.10.6:6789:/       /mnt/cephfs     ceph    name=vmuser,secretfile=/etc/ceph/vmuser.secret,noatime,_netdev  0       2

When I mount -a (as root), I keep getting a "mount error 110 = Connection timed out". I can ping all the monitors, so I know this VM can reach them. For some reason, they just aren't responding with anything the VM can use.

Interestingly, if I "cat" /etc/pve/priv/ceph.mon.keyring on the host, it only shows an Admin user.

Does anyone have any tips I could try to get this working???? I'm pulling out what little hair I have left.