Hello all,
Just wanted to issue a heads up to anyone using synology NFS exports as proxmox storage! Synology is starting to support BTRFS volumes with some advanced features on their synology devices. However, proxmox seems to have an issue with that. For example when I'm restoring a vzdumped VM to said proxmox nfs/btrfs storage most of the times it fails (not immediately but could be at any point during restore) with:
** (process:3924): ERROR **: restore failed - bdrv_write to failed (-116)
/bin/bash: line 1: 3923 Broken pipe lzop -d -c /mnt/pve/ioSafe/dump/vzdump-qemu-0000-Template-Win7-32bit-Syspreped-2014_08_26-12_27_04.vma.lzo
3924 Trace/breakpoint trap | vma extract -v -r /var/tmp/vzdumptmp3921.fifo - /var/tmp/vzdumptmp3921
temporary volume 'VM:101/vm-101-disk-1.qcow2' sucessfuly removed
TASK ERROR: command 'lzop -d -c /mnt/pve/ioSafe/dump/vzdump-qemu-0000-Template-Win7-32bit-Syspreped-2014_08_26-12_27_04.vma.lzo|vma extract -v -r /var/tmp/vzdumptmp3921.fifo - /var/tmp/vzdumptmp3921' failed: exit code 133
Also running VMs using that storage appear to lose access to their disk giving IO errors.
The odd thing is that I also tried copying many many gigs of large files from the proxmox host command line to the same NFS export and then did MD5 checksum comparison. I got no errors during the copy and all checksums matched. So besides proxmox VM restore to this export and live VM using a cqow2 file on this export I was unable to duplicate the issue by other methods.
This seems to be a problem only on write, not read. So if I restore a VM from vzdump files on the NFS share to some other storage I don't see the problem.
If I remove the BTRFS volume on the syno and create the volume on the Synology as EXT4 and export with NFS, I have absolutely no problems.
My test setup was with HP ProLiant servers connected via a 1Gbit/sec network to a Synology DS716+ configured with Single volume per disk group formatted as BTRFS and exporting shares with NFS. I have not tried iSCSI.
Just wanted to give you a headsup in case someone hits the same problem. BTRFS is new and just starting to be deployed with Synology devices so I'm sure others will have this issue as well.
Also, if anyone has any ideas as to what could be causing it, let me know!
Just wanted to issue a heads up to anyone using synology NFS exports as proxmox storage! Synology is starting to support BTRFS volumes with some advanced features on their synology devices. However, proxmox seems to have an issue with that. For example when I'm restoring a vzdumped VM to said proxmox nfs/btrfs storage most of the times it fails (not immediately but could be at any point during restore) with:
** (process:3924): ERROR **: restore failed - bdrv_write to failed (-116)
/bin/bash: line 1: 3923 Broken pipe lzop -d -c /mnt/pve/ioSafe/dump/vzdump-qemu-0000-Template-Win7-32bit-Syspreped-2014_08_26-12_27_04.vma.lzo
3924 Trace/breakpoint trap | vma extract -v -r /var/tmp/vzdumptmp3921.fifo - /var/tmp/vzdumptmp3921
temporary volume 'VM:101/vm-101-disk-1.qcow2' sucessfuly removed
TASK ERROR: command 'lzop -d -c /mnt/pve/ioSafe/dump/vzdump-qemu-0000-Template-Win7-32bit-Syspreped-2014_08_26-12_27_04.vma.lzo|vma extract -v -r /var/tmp/vzdumptmp3921.fifo - /var/tmp/vzdumptmp3921' failed: exit code 133
Also running VMs using that storage appear to lose access to their disk giving IO errors.
The odd thing is that I also tried copying many many gigs of large files from the proxmox host command line to the same NFS export and then did MD5 checksum comparison. I got no errors during the copy and all checksums matched. So besides proxmox VM restore to this export and live VM using a cqow2 file on this export I was unable to duplicate the issue by other methods.
This seems to be a problem only on write, not read. So if I restore a VM from vzdump files on the NFS share to some other storage I don't see the problem.
If I remove the BTRFS volume on the syno and create the volume on the Synology as EXT4 and export with NFS, I have absolutely no problems.
My test setup was with HP ProLiant servers connected via a 1Gbit/sec network to a Synology DS716+ configured with Single volume per disk group formatted as BTRFS and exporting shares with NFS. I have not tried iSCSI.
Just wanted to give you a headsup in case someone hits the same problem. BTRFS is new and just starting to be deployed with Synology devices so I'm sure others will have this issue as well.
Also, if anyone has any ideas as to what could be causing it, let me know!
