gleaming wharf
Hello,
I would like to know it this is possible to set a snapvault, snapmirror, SVM DR relation directly to the storagegrid, to tier all the snapvault data in storagegrid already at the creation of the relation ?
the reason is that the space of the local destination aggregate is limited and I would like to bypass the stockage of the baseline in the local aggregate.
Thank you !
gleaming wharf
I found the answer, the destination of the relation will take the same tiering policy as the source
worldly shoal
The policies do not have to be same for a snapvault. You can use an auto policy on the source and an all policy on the destination. We have different settings for tiering-minimum-cooling-days on our destination than we do on the source.
worldly shoal
This is one of my regular commands (in diag mode) to monitor the footprint on my destination:
vol show-footprint -aggregate !aggr0* -field aggregate,volume-blocks-footprint-bin0 -sort-by volume-blocks-footprint-bin0 -volume-blocks-footprint-bin0 >2t
I've been known to do a vol move from one aggregate to another and setting tiering-policy all on the vol move to force it to tier, and then revert the tiering policy back when it's done.
Note that small filers with SATA drives are lousy destinations for SV with FabricPool tiering - they simply don't have the I/O bandwidth to do the tiering calculations and you become spindle-bound with no way out.