FabricPool: how cold data will be flowed into the other type of composite aggregate?

[netappmagic]

Two aggregates here:

one is attaching to AWS-S3, aggr-aws, the other is attached to StorageGrid, aggr-sg.

Suppose a volume in aggr-aws with "snapshot-only" policy enabled, and cold data was already in the cloud tier, when I "volume move" this volume to aggr-sg:

1.   Will cold data be pulled  back to the performance tier first?
2.   If yes, how long will cold data be staying in the performance tier, before get moved to aggr-sg?

[AvivDeg]

Hi,

The data shouldn't be kept on the performance tier when doing vol move.

If it's 9.6 and you had the same bucket for all the aggrs, ONTAP would have moved just the hot blocks to the other aggr.

[netappmagic]

@AvivDeg  

What if aggrs are not in the same bucket?

As I said, Source aggr is attaching to AWS S3, Destination is attaching to StorageGrid?

Thanks,

[AvivDeg]

The tiered snapshots data shouldn't be kept on the performance tier when doing vol move

[netappmagic]

@AvivDeg 

Due to the fact that the cold snapshots are currently in the cloud, are you saying that the vol move will migrate data from the cloud directly to StorageGrid? If yes, I would assume the data path should at least pass through the performance tier if not be staying on the performance tier for too long , right?

[AvivDeg]

The path is passthrough ONTAP. ONTAP shouldn't keep the data on the performance tier, just the LIFs/CPU etc will work a bit harder.

[netappmagic]

Thanks for the clarification. 

One more follow-up. Can you tell which LIF on the node the data will passthrough, is it data lif's or Admin lif?

[AvivDeg]

I'm guessing it will be the intercluster LIFs