FabricPool and hardware lifecycle management

[LORENZO_CONTI]

Hello,
I have a question regarding using FabricPool aggregates and performing replacement for nodes in a cluster (for instance in case of a tech refresh). If I understand, starting from Ontap 9.6 we can benefit from optimized volume move, in case source and destination aggregates are using the same object store.

In that case, when performing the hardware refresh we will:
1) create a fabricpool on the new node(s) using the same object store of the ones to be decommissioned
2) perform vol move (with tiering policy "auto", only hot data will be moved)
3) destroying old aggregate(s) 
4) cluster unjoin of the old nodes

Do it will work this way? 😉

Thanks

Lorenzo

[Fabian1993 · Accepted solution]

Hi @LORENZO_CONTI,

absolutely right, starting with 9.6 if a volume move’s destination aggregate uses the same bucket as the source
aggregate, data on the source volume that is stored in the bucket does not move back to the performance
tier.

Your Plan makes sense 🙂

[Mjizzini]

However there are few conditions that can make the volume move to be unoptimized, where all the data in cloud tier will be copied to performance tier and then copied back to cloud tier based on the tiering policy.

Why does volume move copy cloud tiered data on releases 9.6 and above

[tahmad]

Hello @LORENZO_CONTI ,

Did you find the answer for your question?

[LORENZO_CONTI]

Hello @tahmad , basically yes, but I would like to test an optimize vol move as soon as possible 😉