Correct, Isilon's SmartDedupe functionality is fixed-length block deduplication of the data at-rest.

Inline? 

Why would you ever want it?  I ask, because the logic is different with Isilon than most scale-out NAS platforms where this kind of thing is offered now and then. 

Scale-up logic for dedupe

Think of it this way, in a scale-up NAS, you have small filesystems, usually with 16 or 64TB max sizes, (of course that's changing some).  Those small filesystems though have limits on their capacity which is the point, and when they grow they consume space from some underlying pool which is comprised, itself of raid groups.  So by temporarily using disk space, if the volume or filesystem has to grow, you've eaten up something real, disk space on the pool backing the volume. 

Scale-out logic for dedupe

But with Isilon?  The whole cluster is a single gigantic filesystem, and that could be 100TB or 50+PB.  So using a bit more space temporarily and then shrinking the data down after the fact as a low-priority job, in the background, is far more efficient than trying to waste CPU cycles or RAM and slow down an incoming write to dedupe the data before it's written.

Make sense?  The only value I could see here would be for an RFP check-box, when someone demands the feature even though they don't know why they really want it.

For any roadmap discussions (to which I have no insight as a DellEMC Select partner), of course contact your DellEMC account team.  And please if you can think of a reason why this would be valuable outside of what I described, pass it along, and ask them to submit an RFE (request for enhancement).

Chris Klosterman

Principal Pre-Sales Consultant, Datadobi

chris.klosterman@datadobi.com

To add to what Chris already stated, here is the white paper on SmartDedupe
https://www.emc.com/collateral/white-paper/h12395-isilon-smartdedupe-wp.pdf

Here is the original technical demo published by Dell EMC:
Technical Demo: EMC Isilon SmartDedupe - YouTube