We have been running a Proxmox 1.7 installation with the latest kernel with 3 (test) KVMs for a few weeks. The nightly backups (0:15 each night) were running onto an NFS share on a standalone CentOS host. This worked well for a few weeks.
We then virtualised that CentOS (fresh install of a KVM under Proxmox, restore into that new VM) and it is running now under Proxmox. We did not recreate the (Proxmox backup) NFS share on the newly build CentOS VM
At that time I relalised that we had forgotten to remove Proxmox' link to the non-existing NFS share. I removed that link through Proxmox web interface, disabled the backup jobs and then even deleted the backup job.
Last night Proxmox stopped with a kernel panic at 0:15 with the following error:
Feb 23 00:15:51 twt250sv kernel: nfsd D ffff8801ba315000 0 2908 2 0x00000000
Feb 23 00:15:51 twt250sv kernel: ffffffff81491c30 0000000000000046 0000000000000000 00000002a5c06540
Feb 23 00:15:51 twt250sv kernel: ffff8801b85a9460 ffffffff00000000 000000000000fa40 ffff8801b85a9fd8
Feb 23 00:15:51 twt250sv kernel: 0000000000016940 0000000000016940 ffff8801ba315000 ffff8801ba3152f8
Feb 23 00:15:51 twt250sv kernel: Call Trace:
Feb 23 00:15:51 twt250sv kernel: [<ffffffff810b6d99>] ? sync_page+0x0/0x46
Feb 23 00:15:51 twt250sv kernel: [<ffffffff81313d95>] ? io_schedule+0x9b/0xfc
Feb 23 00:15:51 twt250sv kernel: [<ffffffff810b6dda>] ? sync_page+0x41/0x46
Feb 23 00:15:51 twt250sv kernel: [<ffffffff813142bf>] ? __wait_on_bit+0x41/0x70
Feb 23 00:15:51 twt250sv kernel: [<ffffffff810b6f5e>] ? wait_on_page_bit+0x6b/0x71
Feb 23 00:15:51 twt250sv kernel: [<ffffffff81066960>] ? wake_bit_function+0x0/0x23
.. snip ..
It looks like there are still some pointers/links to the non-existing NFS share but I don't know where these would be?
Michael
We then virtualised that CentOS (fresh install of a KVM under Proxmox, restore into that new VM) and it is running now under Proxmox. We did not recreate the (Proxmox backup) NFS share on the newly build CentOS VM
At that time I relalised that we had forgotten to remove Proxmox' link to the non-existing NFS share. I removed that link through Proxmox web interface, disabled the backup jobs and then even deleted the backup job.
Last night Proxmox stopped with a kernel panic at 0:15 with the following error:
Feb 23 00:15:51 twt250sv kernel: nfsd D ffff8801ba315000 0 2908 2 0x00000000
Feb 23 00:15:51 twt250sv kernel: ffffffff81491c30 0000000000000046 0000000000000000 00000002a5c06540
Feb 23 00:15:51 twt250sv kernel: ffff8801b85a9460 ffffffff00000000 000000000000fa40 ffff8801b85a9fd8
Feb 23 00:15:51 twt250sv kernel: 0000000000016940 0000000000016940 ffff8801ba315000 ffff8801ba3152f8
Feb 23 00:15:51 twt250sv kernel: Call Trace:
Feb 23 00:15:51 twt250sv kernel: [<ffffffff810b6d99>] ? sync_page+0x0/0x46
Feb 23 00:15:51 twt250sv kernel: [<ffffffff81313d95>] ? io_schedule+0x9b/0xfc
Feb 23 00:15:51 twt250sv kernel: [<ffffffff810b6dda>] ? sync_page+0x41/0x46
Feb 23 00:15:51 twt250sv kernel: [<ffffffff813142bf>] ? __wait_on_bit+0x41/0x70
Feb 23 00:15:51 twt250sv kernel: [<ffffffff810b6f5e>] ? wait_on_page_bit+0x6b/0x71
Feb 23 00:15:51 twt250sv kernel: [<ffffffff81066960>] ? wake_bit_function+0x0/0x23
.. snip ..
It looks like there are still some pointers/links to the non-existing NFS share but I don't know where these would be?
Michael