I have run into an issue a couple of times in that guest OS's slow down dramatically if the PBS server doesn't perform for whatever reason.
Previously I had a network issue, which prevented backups from being written at a reasonable speed and it caused the guest machines being backed up to slow down to the point of becoming unusable. It was pointed out to me that the availability of the PBS is crucial in the backup process, which makes sense, but the negative effect on the guest OS doesn't make sense.
I backup to a remote proxmox cluster. When there is any problem (networks, PBX server, disk issues, etc) it affects the guest OS being backed up dramatically. This is a serious problem. If a machine being backed up becomes unusable, then it indicates a design problem in my mind, but may there's a priority tuning that can fix this?
A backup is made from a snapshot, so once the snapshot has been taken, essentially the diff of the existing backups and the new one is taken and the data sent to the PBS server., correct? However, even though I have a guest OS with 16 cores and 32GB of RAM, when there is any issue with the data transfer, the guest OS becomes so slow that it can be considered unusable. Surely this cannot be considered normal?
How can I address this? Production machines should never slow down dramatically because a backup is being made, not so?
Previously I had a network issue, which prevented backups from being written at a reasonable speed and it caused the guest machines being backed up to slow down to the point of becoming unusable. It was pointed out to me that the availability of the PBS is crucial in the backup process, which makes sense, but the negative effect on the guest OS doesn't make sense.
I backup to a remote proxmox cluster. When there is any problem (networks, PBX server, disk issues, etc) it affects the guest OS being backed up dramatically. This is a serious problem. If a machine being backed up becomes unusable, then it indicates a design problem in my mind, but may there's a priority tuning that can fix this?
A backup is made from a snapshot, so once the snapshot has been taken, essentially the diff of the existing backups and the new one is taken and the data sent to the PBS server., correct? However, even though I have a guest OS with 16 cores and 32GB of RAM, when there is any issue with the data transfer, the guest OS becomes so slow that it can be considered unusable. Surely this cannot be considered normal?
How can I address this? Production machines should never slow down dramatically because a backup is being made, not so?
