INFO: resuming VM again after XXX seconds (backup mode stop)

[alexunderboots]

Good day. please help me
How to increase the time between the start of a full backup and the start of the VM.
Or start the VM after "backup mode stop"

PVE 8.3

 

[fiona]

Hi,
why would you like for the VM to be blocked/stopped longer? The backup procedure keeps track of changed blocks to guarantee a consistent state of the disks (and you should install/enable the guest agent to guarantee a consistent file system state). To install the block tracking and start the backup, only a short time is needed, which is why the VM is able to resume right afterwards. So there is simply no need to wait longer.

 

[alexunderboots]

Hi,
why would you like for the VM to be blocked/stopped longer? The backup procedure keeps track of changed blocks to guarantee a consistent state of the disks (and you should install/enable the guest agent to guarantee a consistent file system state). To install the block tracking and start the backup, only a short time is needed, which is why the VM is able to resume right afterwards. So there is simply no need to wait longer.

guest agent installed, but some small VM (40Gb, specially mail-servers MIAB) not correct start during "backup mode stop" (mail base mounted by NFS)

 

[fiona]

guest agent installed, but some small VM (40Gb, specially mail-servers MIAB) not correct start during "backup mode stop" (mail base mounted by NFS)

If you have guest agent installed, (by default) the filesystem is frozen during the time the backup is started to ensure consistency. For some services you might need to configure an fsfreeze hook, but otherwise there shouldn't be much advantage using stop mode instead of snapshot mode.

What exactly is the failure during start? How does the network/CPU/IO load look like around the time of the issue? If it's because of the NFS, is that on the same network as backup? Maybe it's enough to configure a bandwidth limit for backup.