Would it be good to have some progress bar for running backups? Right now I've got a CT showing "created new dynamic index 2" for 20+ hours....
Backup progress percent or bar or else?
- Thread starter grin
- Start date
-
- Tags
- backup user interface wishlist
On the server side? You cannot really know how much and over what time spans the client will send data...
On the client side it would be nice to have some more verbosity, a definitive percent indicator is also here not to cheap - we would need to either scan all files first (really slow) or at least get the disk usage first (which is also in the same range of slow).
It is sadly just not as easy as with VMs, where we can track the relevant information with QEMU easily.
On the client side it would be nice to have some more verbosity, a definitive percent indicator is also here not to cheap - we would need to either scan all files first (really slow) or at least get the disk usage first (which is also in the same range of slow).
It is sadly just not as easy as with VMs, where we can track the relevant information with QEMU easily.
Apologies if I should have detailed what I meant. Let me:
There is two side of pbs+pve backup process: the pve sender and the pbs receiver side.
On the sender side the total volume may be known, especially if vzdump made a snapshot, so percentages ("progress") can be calculated.
I don't specifically know which subsystem does the actual sending but either the vzdump part or the backup-client part has the possibility to output the progress; yet I do not mean percents at all costs, bytes transferred (or even bytes processed) would belp to see what's moving and at what speeds.
On the receiver side it is true that we don't (or might not) know the total but still it would be very helpful to see the progress in bytes, like the amount of transferred bytes (/megabytes or whatever magnitude we are at) every 10 seconds, or minute or so. It would help to see how the backup goes, especially if I am already aware of the total size based on previous backups.
Both may also report the transfer speed observed, along the progress.
In case of VM there's also the rust process transferring the data which may be able to report the progress.
There is two side of pbs+pve backup process: the pve sender and the pbs receiver side.
On the sender side the total volume may be known, especially if vzdump made a snapshot, so percentages ("progress") can be calculated.
I don't specifically know which subsystem does the actual sending but either the vzdump part or the backup-client part has the possibility to output the progress; yet I do not mean percents at all costs, bytes transferred (or even bytes processed) would belp to see what's moving and at what speeds.
On the receiver side it is true that we don't (or might not) know the total but still it would be very helpful to see the progress in bytes, like the amount of transferred bytes (/megabytes or whatever magnitude we are at) every 10 seconds, or minute or so. It would help to see how the backup goes, especially if I am already aware of the total size based on previous backups.
Both may also report the transfer speed observed, along the progress.
In case of VM there's also the rust process transferring the data which may be able to report the progress.
Very very good idea to show the pve vzdump progress in the GUI
I think is a simple thing as this would be something to pipe the data to a subsystem that can show it in the Sender Side. The backup side really doesnt matter as if this feature is implemented in the sender side it would always work no matter wich technology is ran in the reciever side
I think is a simple thing as this would be something to pipe the data to a subsystem that can show it in the Sender Side. The backup side really doesnt matter as if this feature is implemented in the sender side it would always work no matter wich technology is ran in the reciever side
This thread is over a year old, but it's worth ressurecting. I recently had to move data to a disaster recovery site and for 3 days I have been in total darkness as to what is happening.
proxmox-backup-client needs some information like rsync provides. -vP -v would show the file it is currently backing/restoring and -P would show progress for that file
sync jobs - since the server knows the sizes (index.json.blob) some form of % should be possible. At least for the current VM. Having a status window that simply repeats GET chunk XXXXX is pretty useless
proxmox-backup-client needs some information like rsync provides. -vP -v would show the file it is currently backing/restoring and -P would show progress for that file
sync jobs - since the server knows the sizes (index.json.blob) some form of % should be possible. At least for the current VM. Having a status window that simply repeats GET chunk XXXXX is pretty useless
Im now in the same boat
Im migrating my Server, means im backupping vm's/containers simply to an usb drive...
It's an usb3.2 ssd, but its anyway slow as hell...
However on vm's luckily im getting progress, but for containers i don't
So im sitting and waiting since hours and came to the idea to google if i can somewhere see the progress of my docker container which is around 1,4tb big
So my Big and loved Proxmox Team, please add that to your to-do list
Cheers
Im migrating my Server, means im backupping vm's/containers simply to an usb drive...
It's an usb3.2 ssd, but its anyway slow as hell...
However on vm's luckily im getting progress, but for containers i don't
So im sitting and waiting since hours and came to the idea to google if i can somewhere see the progress of my docker container which is around 1,4tb big
So my Big and loved Proxmox Team, please add that to your to-do list
Cheers
Realizing that getting the total data size is expensive, I think it would be immensely useful just to have the number of bytes read/written/received/sent. This is already done in one form or another in the PVE task logs. I'd like to see this information also coming from proxmox-backup-client.
Is there a bugzilla feature request yet? Have only found a somewhat related one [1].
If we expect it to be implemented at some point, there should definitely be a feature request in bugzilla.
[1] https://bugzilla.proxmox.com/show_bug.cgi?id=4946
If we expect it to be implemented at some point, there should definitely be a feature request in bugzilla.
[1] https://bugzilla.proxmox.com/show_bug.cgi?id=4946