Data Optimization and FlexGroups - Volume Constituants

[JFM]

Hi,

On an AFF unit, using a flexgroup where constituents are spread across both nodes of the controller, are blocs dedupped across all constituents sitting on the same aggregate only or across all constituents across all aggregate? According to this: https://docs.netapp.com/us-en/ontap/flexgroup/enable-storage-efficiency-task.html I would extrapolate to think that data optimization "domain" is still local aggregates only. Am I correct?

In other words, two identical blocs stored in constituents sitting on two different aggregates wouldn't be optimized and stay as is. 

Thanks for your input. 

[SpindleNinja · Accepted solution]

It's per aggr. 

[JFM]

So, I guess my understanding was right. 

[parisi]

Eh, it's per aggr for cross-volume dedupe.

It's per member/constituent volume for other efficiencies. Check TR-4571 for details.

https://www.netapp.com/us/media/tr-4571.pdf - starting on page 71

[SpindleNinja]

Technically correct, the best kind of correct.