Search results

This server is only for backup for both zfs snapshot and normal backups. So I think it is ok.

So there is nothing wrong and I can do here?

root@backup:~# zpool list NAME SIZE ALLOC FREE CKPOINT EXPANDSZ FRAG CAP DEDUP HEALTH ALTROOT D2D3 7.25T 2.74T 4.51T - - 17% 37% 1.00x ONLINE - D4 3.62T 834G 2.81T - - 15% 22% 1.00x ONLINE - rpool 3.62T 3.15T...

I have a 4 TB Disk and mounted as zfs. I also created an other one with command: zfs create D4/BACKUPS -o mountpoint=/D4BACKUP to store disk backups. But their sizes are different. Please see screen shot. Also their used spaces are also different. Probably because it has zfs snapshot and disk...

There is nothing can be done for already moved one right? It will make a full sync with next run.

I am a bit confused. You mean I HAVE TO have a common (at least one) snapshot after I move a vm to new node TO STOP doing full sync with next pve-zsync run ? Also how about files in /var/lib/pve-zsync/ directory? Are they in any way in subject for before running command to check snapshots?

Ok how about ignoring existing snapshots and let it create a new snapshot and copy it over. It wont take long anyway. Can I delete all existing snapshot with zfs list -t snapshot -o name | grep vm-124| xargs -n 1 zfs destroy -vr along with files in /var/lib/pve-zsync/* ?

Ok I install a new node and move kvm with command: pve-zsync sync --source 1.2.3.4:124 --dest D2 --verbose After the move, on backup server (not a pbs), I have a cron with content: /usr/sbin/pve-zsync sync --source 1.2.3.4:124 --dest D2D3 --name 110 --maxsnap 20 --method ssh; I then update...

When I move a vm to a new node, pve-zsync starts for a new complete snapshot instead of doing only snapshot to backup server (not proxmox backup server, normal proxmox node). And this makes lots of traffic. Is it possible to avoid that?

Yes api or a simple bash would do the job. Thank you.

I am managing many directadmin and windows mail servers with many domains. I played with PMG. I saw that I have to add my domains to PMG. Is there a way to define ip instead of domains. That would be much easier and I would not have to add each newly created domain that I added to mail servers...

Thank you. Would it be ok for now to use 3rd party like https://github.com/jimsalterjrs/sanoid

I need to use something like pve-zsync sync --source 1.2.3.4:ALL --dest D4 --name ALL --maxsnap 20 --method ssh Or any alternatife script for zfs snapshot with support of all vms in node.

Ok somehow it is fixed now. Probably at isp level issue and they fixed. Thank you for suggestion.

Suddenly, all vm to an other pve stop connecting on port 25. I have no firewall blocking port25, isps also checked against port blocking. But we are unable to find issue. Server-a can connect to server-b via telnet ... 25 but not server-b. I am not using any nat. What else you think I could...

OK ok calm down. I just needed to add hwaddress to interface. All are up now :))

Sorry can not copy via kvm: Any suggestion?

I have install pve on a SATA disk and use vm on ssd disks. In case of os fails, so I can simply reinstall pve. My question is, would sata disk as os disk cause performance issues?

I am getting same thing with my 2 days old ssds.

No it is on shared mode. I think I have to use different ip.