Hello,
Our physical Windows 2016 servers are making backups with Windows Server Backup to an iScsi connected disk.
This way we can store our backups on another location and have the benefits of incremental backups, which would not have worked over an regular network share.
During a disaster recovery we would then connect the iScsi disk to another server.
On the server we use "VSSADMIN LIST SHADOWS" to find the shadowcopy of the date that we want to go back to.
Then using "MKLINK /d <symbolic-link> <target>" we make that shadowcopy available for usage. Within the file explorer you can see the link and also view de WSB data, with the modified date of that specific point in time.
On this symbolic link we create a share so the system that needs te be restored can be pointed to get the WSB data from there.
This works and all is fine.. So far so good.. but now the problem we encountered.
That was in de WSB data from a point in time perspective.
One DR test worked without any problems and we could restore the system by using a shadowcopy from let's say two days back in time.
On another server we used the exact same procedure but the restore of two days back in time failed and gave the error:
"System image restore fialed: The mounted backup volume is inaccessible. Please retry the operation (0x8078004F)"
By trial and error I found out that the first WSB data back in time that was usable was from 40 days ago. All other shadowcopies untill the one that was made last night failed and could not be used.
Now we have some questions which we haven't found an answer to so far..
Q: Most important.. Is there any way that we can perform a health check of WSB data from our other servers that we haven't checked yet?
Q: Does anyone have a possible answer to why all WSB data after a date is not useable anymore. Personally I have a gut feeling that is has something to do with all being differentials.
Q: Can we limit the number of days of unuseable data by forcing a full backup once a week and the other days differential.