Justin Parisi

I know I'm seeing a lot of questions about it, but I'm the flex group team here, so I should be seeing those questions.

But as someone who works in migration services, how are you seeing the uptick with that?

Yeah, and I mean, you talked about large data structures, you know, over 100 terabytes and that sort of thing.

It's not just for that.

I mean, it's for things that are smaller.

You can split up those workloads across multiple FlexVols and get advantages for performance on, you know, volumes as small as 20 terabytes, right?

So you can definitely split it up that way.

Yeah, it basically takes advantage of the CPU affinities on the nodes, so you're able to spread that out across multiple affinities.

With a FlexVol, when you're doing metadata operations, you're basically single-threaded to a CPU.

So you want to try to spread that out as much as possible.

Then you get more efficient use of your hardware.

And then when you start scaling out across nodes, then you can start having a true scale-out file system as opposed to a single-node entity like a FlexVol.

All right, Adrian, thanks so much for joining us and talking about migration.

We covered a lot of ground here.

We talked about NAS, we talked about SAN, we talked about XCP, Flex Groups.

I mean, what more could you ask for?

So I actually have it written in there.

I have to say it with a certain exuberance, and I get royalty.

Nice.

Of $0 per mention.