Double DeDupe for Aux Copy?

Hi @Sudharshan Kumar ,

DASH Copy jobs transmit only unique data blocks to the destination, which reduces the volume and time for an Auxiliary Copy job by up to 90%. - This is ideal for low bandwidth connections also.

More details on Commvault Deduplication can be found here: https://documentation.commvault.com/commvault/v11/article?p=12401.htm

If you require any specifics on this, please let us know.

Best Regards,

Michael

Hi @MichaelCapon : 

Thanks for your reply, interested to know why we doing Dash Copy for separate deduplication? already in Primary Copy Deduplicated stored in our DiskLib. 

Hi @Sudharshan Kumar ,

The Primary Copy of the policy is the Primary location of the Deduplicated Data and will be referenced by the DDB for that Policy/Storage Pool associated.

DASH Copy would be used to replicate Deduplicated Data to another Library in Deduplicated format, typically this would be in another location and attached to another MediaAgent.
- Due to the resources being separate, those will need a DDB in the form of a Global Deduplication Policy/Storage Pool configuration in order to Deduplicate data.

DASH Copy process would effectively then check source side signatures (In Source DDB) against the destination DDB’s signatures, any unique blocks identified would then be replicated across.

I hope this makes sense?

Best Regards,

Michael

It is still confusing as the primary copy is deduped, and the ddb for it is created and known, why not just move the deduped chunks to the remote MA, as the DDB is not needed for restores anyway, wouldnt this save a lot time chewed up by the aux copy ddb process ?

The DASH copy is re-deduped.  Keep in mind, you might have much longer retention on the Aux Copy than the Primary (and most likely do) so it’s not as if the DDB from the Primary will aid the Aux Copy in determining what can be deduped and what can’t be.

You’re 100% right that the DDB is not needed for restores; it’s only needed to dedupe the written blocks, per copy uniquely.

To add to Mike’s comments, every Aux Copy is another standalone copy of the data. This is done for the exact same reason that “replication” is not “data protection”.

What if you mistakenly delete a client or job from the primary backups? If we’re replicating blocks, the client or job is removed from both. With the Aux Copies being standalone copies, the data still exists and can be recovered back to the primary.

Thanks,
Scott