Migrate storage repository back to same host - OS Upgrade

[L33T]

Hi,

We are running B&R 9.5 SP1, we have 10.9TB of backup files (reverse incremental) in a repo but wish to upgrade the Windows OS on the backup host. What methods can we use to ensure we keep the current backup files as we do not wish to loose the last 28 days backups for restore purposes. As they are of such a size copying them across the network would make them out of date by the time we could get that off and back on again.

Any help would be great.

Cheers,

[foggy]

Are those backups residing on the backup server itself (Windows type repository)? Any chance you could unmount the disk containing them prior the upgrade and mount it then back?

[L33T]

Hi,

They are on the backup host local storage and the logical array needs to be erased and re-presented back to the host. We have an external array that we are to merge with the internal (HP SAS) but either way i`m not sure how to keep this retention without copying it somewhere else. Would the only option be to get another repo host ready (hp microserver with 4x6TB RAID10, copy the existing backup files to that, setup the backup host and then use the copied repo as an additional repo for restoration purposes?

Just not sure what`s the best bet with this.

cheers,

[foggy]

If you need to re-format the disk containing backups, you need to temporarily copy them somewhere, I do not see other options.

[L33T]

Yep same here, i`m going to perform the following:

Order a hp microserver with 4x6tb disks and setup as RAID-10
Copy the current files to it
Setup the new veeam backup server install on the current hardware merging both the internal external arrays, format as 64k blocks and use the /L large volume switch
point to new repo
point to old repo (i presume then i can use it for historical restores e.g. older copied backups
move on forwards with the new repo

Any other steps i should perform?

Cheers,

[foggy]

The outlined steps should work.

[L33T]

thanks, i`ll post progress asap

Cheers,

[L33T]

additional thoughts on this...

If i create a scale out repository and pick data locality can i select my existing repo (with the last 30 days of backups stored on it) as an extent and have that then spread across both extends, after that rebuild the veeam host and bring back into the scale out pool keeping all the data reducing the downtime and loss of days?

regards,

[foggy]

Do you mean using evacuate option to migrate data? This will also take time comparable with the file copy to some third location.

[L33T]

Hi.

ok what does:

Evacuate actually do
Data locality, what happens if you add an existing standard repo with data in it to a scale out repo picking this option?

I guess what i`m saying is can i utilise these features to retain the 30 days data already in a standard repo, get the backup server re-installed then continue to use it / copy it back to the fresh new backup host if i add it back to the scale out repo afterwards?

Cheers,

[foggy]

If you add existing repository as an extent to the scale-out one and then click Evacuate, data will be moved to other extents.

[L33T]

ok cool, would the existing backup data be affected / lost in the repo i wish to convert as an extent? If not i`ll just go that route as i would like to use scale out when this migration is complete.

[L33T]

Forgot to add, when this has been done will veeam then automatically distribute the backup files back to the local extent (the one that existed before the re-install) when it has been added as an extent to the SOBR?

[foggy]

Existing data will not be affected. Data will not be re-distributed on adding an extent, but will continue to be written according to the selected policy.

[L33T]

This all worked fine Job sorted.