deduplication between different jobs

[mcz]

Hi everybody,

I've read in this thread that there's no mapping between old and new backup possible. Now we are copying backups via BCJ to a SOBR which has a capacity tier extend (object storage) added. Now what happens if a new BCJ creates a new chain, will offload do deduplication because backed up vm's are identical or will it start from scratch?

Thanks for the clarification!

[HannesK]

Hello,
I moved your question to the object storage forum.

the offload deduplication for object storage works very similar to REFS block cloning deduplication. That means that only blocks in the same backup chain are deduplicated.

So the answer to your question is: no

Best regards,
Hannes

[mcz]

Hi Hannes,

thanks for your response. I could imagine why the decision has been made but finally it's a big stumbling block from time to time. For instance, the new copy mode for a BCJ in v10 can only be used when creating a new job which then creates a new chain and would offload to a new folder - so no deduplication would take place. Now since we're using GFS jobs, you would need a lot of space on s3 to start with a new chain to regain that deduplication benefit that you had before.

Any plans to extent deduplication functionalities across multiple chains on the same bucket in the future?

[HannesK]

that would be a very complex task...I would not wait for that...