Skip to main content

size of archive does not shrink when decreasing number of backups to keep

Thread needs solution

I was running low on diskspace (still 60G free, but slowly decreasing) on on of my backup destinations, so I thougth to reduce the nuber of backups to keep per VM from 14 to 7 (backups are done daily, so having a backlog of 1 week instead of 2) and only have 2 weeks on the 2nd storage location, but after 2 backups passing by the primary archive is still the same size :-(
Any ideas what could be the cause ?
I'm running vmProtect 8.0 #8148.
When displaying the recovery points I also notice, that the oldest backup of each machine is ~ 240G for *every* VM, even if the VM itself only has 4G of diskspace. To my mind the software is full of bugs, but as we already have paid the money, we have to live with it :-(

Gerhard

0 Users found this helpful

The backup archive files do not shrink when recovery points are removed. The unused space is marked as empty and can be used for future backups. VMware vmdk files behave the same way. One reason that I can think of for this behavior is that to shrink the archive would require a "consolidation" like what happens with VMware snapshots. This would be disk intensive and take a long time, and I have heard of 10-20GB per hour as being common performance. Obviously when deleting say 100GB of recovery points, the archive would be unusuable, and most likely the backup device would have poor performance due to the high disk activity.

Another point - consolidation can require a significant amount of disk scpace, up to the archive size, so in a low free space situation it would not help anyways.

Your best option is to delete the archive and start over (moving it somewhere else if possible, such as a USB drive, by installing the Windows Agent on a workstation and using the dual-destination feature).

thx for your answer. will do as you proposed and delete the archive, as I still do have twice the recovery points hanging around on the dual destination backup place. regarding the size display of recovery points, do you also see this strange behaviour - one very very large value at all the oldest recovery points, and this value nearly the same at all VM's ?

frestogaslorastaswastavewroviwroclolacorashibushurutraciwrubrishabenichikucrijorejenufrilomuwrigaslowrikejawrachosleratiswurelaseriprouobrunoviswosuthitribrepakotritopislivadrauibretisetewrapenuwrapi
Posts: 22
Comments: 3800

Hi Gerhard,

The size of the first recovery point is specifically calculated as full one, i.e. taking into account all the chunks of data inside the archive required to restore the particular VM + VMs backed up at the 1st backup attempt (i.e. total full backup size). This was done due to specifics of single archive structure (where all VMs backups are stored inside one archive file) and situations where 1st backup in chain must be cleaned up due to retention rules - after such clean up the sizes are re-calculated and in order to make them shown up as full ones (to make it easier distinguishing one from another) there was a decision made to set the recovery points sizes in UI to total archive size. This is just a UI sizes which are not reflecting the actual sizes of backups - the actual sizes can be seen in the left location statistics pane (where archives sizes are calculated by fact size of the files, rather than by trying to calculate the data amounts inside the archive structure).

Thank you.
--
Best regards,
Vasily
Acronis vmProtect Program Manager