migrationg storage backend ends in new copy of all data

felipe

Active Member
Oct 28, 2013
221
6
38
hi,

does it makes sense when you do a storage migration to backup the whole disk again?

not only that the first backup is many times slower. i also guess the old disk will stay there on the backup server using data.

example: migrating virtio0 from local disk to ceph means that the backup server will backup the whole disk again fully. but why? its the same disk with the same data...
 

felipe

Active Member
Oct 28, 2013
221
6
38
question to proxmox team:
1) is that itentionally
2) whats with the data of the "old" virtual disk? is it also on the backup data set on the backup server or will it be deleted?
 

dcsapak

Proxmox Staff Member
Staff member
Feb 1, 2016
8,278
1,049
164
34
Vienna
my guess is that on a move disk, the 'dirty-bitmap' gets lost (idk if that will change in the future)
but in any case, even if the backup will read the disk again, there should be no additional network transfer nor data usage on the backup server

the backup first gets the list of chunks from the last backup and only sends those that are missing
 

felipe

Active Member
Oct 28, 2013
221
6
38
if the data usage is double i dont know because it is hard to find.
i will make a check to test if there is no network load. i just saw that it started to read all over again but did not check on the network.
but something to have in mind because it behaves like a new first upload which ist many times slower then a normal backup with dirty bitmap diffs. so when moven a lot of data to another storage the impact on the backup time will be huge. (for example moving 10TB will need some time to read all data again..)

btw: attached nfs storage does not show any network/iops etc. data but local disks do... (or local disks on a pbs vm ceph backed...)
 

dcsapak

Proxmox Staff Member
Staff member
Feb 1, 2016
8,278
1,049
164
34
Vienna
so it works and there is no network activity for the actual backing up?

data but local disks do...
makes sense since pbs has to write logfiles and so on
 

felipe

Active Member
Oct 28, 2013
221
6
38
INFO: resuming VM again
INFO: virtio0: dirty-bitmap status: created new
......
INFO: backup is sparse: 18.52 GiB (46%) total zero data
INFO: backup was done incrementally, reused 37.55 GiB (93%)
INFO: transferred 40.00 GiB in 235 seconds (174.3 MiB/s)

so it looks like it was not tranfered. as this vm was backed up und the nfs share attached to the pbs i could not check for i/o or writes etc....
but i will have in mind that it wil take nearly the same time as a primary backup because of reading the whole disk.
 

About

The Proxmox community has been around for many years and offers help and support for Proxmox VE, Proxmox Backup Server, and Proxmox Mail Gateway.
We think our community is one of the best thanks to people like you!

Get your subscription!

The Proxmox team works very hard to make sure you are running the best software and getting stable updates and security enhancements, as well as quick enterprise support. Tens of thousands of happy customers have a Proxmox subscription. Get your own in 60 seconds.

Buy now!