Hi all,
times ago I used Local Syncmirror to "move" data from old aggregates created with 144 FC disks to new 300 GB FC ones and I remember that the old:new ratio was 2:1, i.e., from an aggregate on 56 144 GB disks on 4 shelves to a new one on 28 300 GB disks on 2 shelves...But I don't remember other stuff related to raid group, geometry of them and so on...simply I used the aggr mirror and Ontap chose the needed disks from a pool of spare (Pool 1 as usual)
Now I've a similar issue but with a lot of doubts (re)reading the manual. First of this is that manual states that from original aggregate and the mirrored one disks must be the same number...and I ask myself how I was able to do that job times ago?!?
Anyway the problem is this:
Current aggregate: 56 disks on 4 shelves of 300 GB but with rg and disk geometry wrong (rg0 20 disks, rg1 26 and rg2 8, a couple of spare). The customer wants to rebuild this aggregate using Local Syncmirror a first time to transfer data to other disks available and a second time to put the data back on the 300 GB disks with a correct rg geometry (only 2 rg with 26 disks each).
I've 8 or more shelves with 144 FC disks available. Can I use Local Syncmirror fm the 4 300 GB shelves to an automatic created mirror on those 8/9 shelves letting Ontap choose the disks from the Pool 1 spares? What's about the aggr mirror xxxx -v "victim_aggrname" found on the man sections and not on the manual? Can I create a new aggregate on the 144 GB disks the use in some way this syntax? Once splitted the mirror I will temporarily have the data on the 144 disks (if work the points before) so I will be able to destroy the 'old', zero the disks and then...? To create another mirror to put the data back on the 300 GB disks will I have to create an aggregate before with the right disk geometry I want and then use this as "victim aggregate"
It's a little bit complicate I know....I apologize in advance for the headaches I will cause in somebody :) but thank you in advance also!
Regards,
PS) we're not discussing here questions about raid group size, geometry, performances and other so I please you to avoid to ask or observe thinks like this. Also, options such as snapmirror, backup and restore and so on useless for the moment because we've about 8 TB of data (long time) to move ahed and back and Local Syncmirror offer the fastest way and overall a way usable "hot" with no disruption
Dott. Giacomo Milazzo [cid:image001.jpg@01CB97F3.CA9D88C0] Technical Account Manager Sinergy SpA Filiale di Roma * 00198. viale Regina Margherita, 269 ' (+39) 3406001045 0644245272 7 (+39) 0226922048 * Giacomo.Milazzo@Sinergy.it
