I have a question regarding the recovery options when a replication pair is in place. I have 3 use cases on our units
1) an NFS store
2) an ddboost store
3) an ddboost Retention Locked store
I have spotted that the DD System manager performs a bunch op operations when setting in place a replication pair. The pain starts if you try to do it manually. The System manager DOESN'T WANT you to create the remote Mtree/ddboost Storage Unit upfront , whereas on the CLI YOU NEED TO
Second of all is that when you perform it on the DD System Manager , this one actually adds the relationships on both nodes. I got puzzled when I did it on the CLI and only had the relationship on one node. Seemingly the repication config is not being propagated to the remote node. Anyhow I limit myself
to the DD system manager (the only task I hold more credit then doing it on the DDOS CLI) ever since.
What I'm puzzled with is :
a ddboost (RL or not) replication pair
I see that the remote Mtree is in place, but the remote node does not see the local storage-unit.
As the remote node does not see the storage unit, I cannot assign a ddboost user on the remote node. So how can I export my ddboost replicationed target ?
Is it at all possible ?
Question 1 :
My practical question is : how should I recover my data loss on prime, using Veeam ,referring back to my ddboost target for recovery purposes ?
This brings me back to the purpose of the replication !
Is its purpose just to sync in one way and if data on source get lost, to recover whilst copying back , period ?
Question 2
Another question is : Suppose I would recover a retention locked ddboost target to the source . Does it operate. What will it actually accomplish ?
Can someone help me to demystify the 2 questions ?
@stefan.timmermans I know it's been several weeks, but this particular community doesn't have very many regular contributors outside of Dell.
A replication target is never directly accessible by backup applications. It's constantly being updated from the source system. The same reason you cannot access a secondary copy using storage replication on a Unity or Powerstore. You have to either break the replication or create a copy. Using fastcopy to another mtree (as you have already done) is the best way to do it on DD. Also, the users may not be the same on both DDs, so when you create the storage unit from the copy, you need to assign a ddboost user.
Then you can add that new storage unit to Veeam as a new repository and rescan the data. The backups will show up under Backups -> Disk (Imported).
I don't know if there will be support from Veeam for being able to control DD replication from within Veeam. It's not really in Dell's interest to help, they would prefer you use PPDM with DD for backup, and you can directly see the secondary copies in PPDM.
It is best to contact either your account manager or professional services to assist you on this. Each customer has different needs and to ensure that we are meeting your needs correctly they can assist you to make sure that nothing is missed.
There is no real customer use case in here only a generic approach on how to recover from a loss of your primary Storage-unit set in a replication context.
I actually came with a solution/workaround, that is I created a storage-unit on the target DD and used fastcopy from the target replica, (which is by design in an Mtree state, despite the source being an Storage-unit) to the newly created storage-unit. Then added this as a ddboost target to veeam for import.
So the operation is basic : how do I recover from my replicated data. I did not ask for guidance on Veeam, just conceptionaly.
I understand that you are using two DDs for one-way replication, transferring backup data acquired from Veeam to a remote location. Regarding question 1, I believe that the Mtree on the replication receiving end is not linked to the STU, which is why it's not being recognized by DDBoost. The receiving Mtree should be read-only, so you should disable replication, enable read/write access, and then link the Mtree and STU via CLI. This should allow Veeam to access it. However, there will likely be a delay when accessing it from a remote Veeam location. I think each site will need a Veeam proxy server. I have a similar configuration implemented in Japan, using two Veeam servers and two DDs.
Yes as the target of a replication is always an Mtree (despite the source being an Storage-unit for a ddboost store at least) and the fact that Veeam needs a Storage Unit, I proceeded as follows for completeness.
1a I created an empty Storage Unit on the Target Unit
1b whilst doing I associated the ddboostuser
2 I used fastcopy to copy the Target Mtree > Empty Storage unit.
3 Add the backuprepository on the Veeam providing used the ddboost user to scroll to the **new Storage unit**
4. Run an Import on Veeam
I believe I halted the replication (instead of removing the relationship)
SteveK821
2 Intern
•
153 Posts
0
1
Posted April 1st, 2026 09:00
@stefan.timmermans I know it's been several weeks, but this particular community doesn't have very many regular contributors outside of Dell.
A replication target is never directly accessible by backup applications. It's constantly being updated from the source system. The same reason you cannot access a secondary copy using storage replication on a Unity or Powerstore. You have to either break the replication or create a copy. Using fastcopy to another mtree (as you have already done) is the best way to do it on DD. Also, the users may not be the same on both DDs, so when you create the storage unit from the copy, you need to assign a ddboost user.
Then you can add that new storage unit to Veeam as a new repository and rescan the data. The backups will show up under Backups -> Disk (Imported).
I don't know if there will be support from Veeam for being able to control DD replication from within Veeam. It's not really in Dell's interest to help, they would prefer you use PPDM with DD for backup, and you can directly see the secondary copies in PPDM.