ripe bridge
Hi all,
when we initially configured SG buckets to use in fabricpool, we adopted the schema "1 bucket per every Ontap aggregate". Now we found this should be an issue if changing the number of performance tier nodes or aggregates. We would prefer having 1 bucket per multiple aggregates.
Is there any way to merge buckets without doing cloud retrieval? I mean not to write blocks on the local tier and then tiering again on the new bucket.
Thanks!
steep peak
Not 100% sure this is what you need but potentially ONTAP FabricPool Mirror might be the way to go, this allows you add 2 cloud tiers to a single local tier. So for each of your local tier aggregates, you add the same 2nd larger SG bucket, and then ONTAP Farbicpool will sync the data from the pre-existing bucket and written to the second bucket. Then later on, you REMOVE the original 1st cloud tier (which also removes the Fabricpool mirror).
I would absolutely run this by your NetApp SE or contact Support.
FarbicPool Mirror is mentioned in our docs and Page 32 of this Techincal Report - https://www.netapp.com/pdf.html?item=/media/17239-tr4598pdf.pdf
On Page 34 there is a section "Enhanced Data mobility" which seems ton confirm my understanding. "Fabicpool Mirror can be used to change the cloud tier attached to a local tier with having perform a volume move".
For reference.
ripe bridge
Thanks Ross, my question is now if the secondary cloud tier must be empty and not used for any other local tiers 🙂
I'll definitely check with SE or support 🙂
steep peak
Good question, I am not sure if you can use the same 2nd clodu tier for multiple local tiers or not. That would be something to check 😄
desert dove
FWIW, we use 1 tenant per grid for FabricPool and then 1 bucket per cluster within that tenant. That's served us well for >3 years.