Data Domain: Write backups if 1 out of 2 Datacenters down

[lukeroom]

Hey togehter,

I need help in order to figure out how we should design our backup infrastructure so that our backups still work even if 1 out of 2 datacenters is not available.
We run 2 DataDomain´s (without ddboost) where we store our backups.

We write the backups on one DataDomain(A) as long as both datacenters are available (normal behaviour). DataDomain(A) replicates all files within this repository to DataDomain(B) which is located in the other datacenter(B).

Right now our problem is: What will happen when DataDomain(A) is not available? For example: DataDomain(A) crashes or datacenter(A) breaks down?
In this case the restore is no problem because we can restore from DataDomain(B).

But how can we continue to write our backups?

We can´t change the repository because we get an error message: (Move all backup files to the new backup repository first).

[foggy]

Lukas, you would need to add the second DD as repository into Veeam B&R, perform rescan and alter all the jobs to point them to existing backups. More details can be found in this thread. Thanks.

[lukeroom]

Hey foggy,

the repository of the second DataDomain(B) is configured read only. Only the DataDomain(A) from the other DataCenter can write to this special directory(share) of DataDomain(B).

regards

[PTide]

Hi Lukas,

I have a question - does DataDomain(B) have a permission to write to that special directory(share) of DataDomain(B)?

Thank you.

[foggy]

So probably I've misunderstood your question: where you're going to continue writing backups in case your primary storage is down? If it is gonna be brand new storage without any backups, then this procedure should help. Otherwise, you can copy backups from the second DD to the new repository and perform the procedure I've referred to above.

[lukeroom]

We can´t copy all files to a new repository. In the repository are round about 30 TB backups.

We write our backups normally on DataDomain(A) and replicate this data with EMC DataDomain mechanism to DataDomain(B) in the other datacenter.
Now there is following case:

DataDomain(A) or the datacenter(A) crashs. The backupjobs can´t therefore write on the DataDomain(A).
Now we want write the backups in a new repository on DataDomain(B) in the datacenter(B).

The procedure to copy all jobs (150) and change the repository is not a good way to solve this problem.

How solve other Veeam / DataDomain Customer solve this case?

[foggy]

The procedure to copy all jobs (150) and change the repository is not a good way to solve this problem.

I believe the entire procedure can be scripted and performed automatically.