Do I only prune once per storage?

Support
#1

I have setup two different Storage ID’s backing up to my Google Drive. So it would look like the following:

Computer 1, 2 and 3 backing up to Storage ID: Gdrive
Computer 4, 5, 6 and 7 backing up to Storage ID: Gdrive_Office

I currently have a Purge Job setup to run nightly for each storage separately. In other words, Gdrive is purged at 2am from Computer 1 and Gdrive_Office is purged at 2:45 from Computer 4.

I have read that I need to be cautious about running a Purge Job from multiple computers against the same storage. Is what I am doing acceptable or is there a better practice?

I was also encountering the VSS_Timeout message even after setting the vss timeout to 600ms. So is the below schedule acceptable?

yALaUUv1

#2

You can prune concurrently.

For more details on how it works, see sections 3.2-3.3 here: duplicacy/duplicacy_paper.pdf at master · gilbertchen/duplicacy · GitHub

This is unrelated to prune

Yes, but you did extra work spreading out backups in time; it is unnecessary. Duplicacy is fully concurrent.

#3

If Duplicacy is fully concurrent than why am I getting the VSS_Timeout messages for the two stations that overlap with the same time?

Note: the image shows the proposed times. Currently, Computer 1 and Computer 4 both start at 7AM and run every 2 hours.

#4

These two things (concurrent backup from multiple machines to a single target and VSS timeouts on any of those machines) don’t have anything to do with each other.

VSS timeout means that Windows Volume Shadow Copy service on your windows host was unable to create a filesystem snapshot within specified time since duplicacy requested it to. This is a windows issue; unsurprisingly, as VSS is notoriously unreliable. You can try repairing your drive with chkdsk c: /f in the elevated CMD. If you are still seeing timeouts after that then you may want to try increasing the timeout value; that is if you can justify the long snapshot taking time (it shall be in fact instantaneous, but since NTFS is not a CoW file system, there is some overhead).

In any case, this has nothing to do with concurrent backup.