This post is more than 5 years old
10 Posts
1
1383
AX4-5 - Hot Spare 'Copy Back'
Hi,
In case a disk fails an array is rebuilt onto the hot spare. Then, when the failed disk is replaced the data is copied from the hot spare to the new disk. Why is the case? Why not just use the new disk as a hot spare, and save yourself a second rebuild process (the way an HP SAN deals with a drive failure).
The only thing I can come up with is that the disk which is set as the original hot spare is not the same spec as the actual array disks, but is this ever the case in reality?
Not sure if I'm missing something here but any clarification is appreciated.
brettesinclair
2 Intern
2 Intern
•
715 Posts
2
June 2nd, 2015 23:00
On arrays running FLARE (on ax/cx and VNX 1) a RAID Group's structure is bound to the Drive ID's of the disks, which is partly made up of the physical location (slot) the drive is in.
On the VNX2 platform, the sparing process works as you've suggested by way of a "Permanent Sparing" process, which makes the spare the permanent member of the group it's replacing, negating the need for a copyback/equalizing process.
Instead of using the physical drive info for ID, the drives are given a Virtual ID made up of their serial number.
AnkitMehta
1.4K Posts
0
June 3rd, 2015 02:00
5 years? It goes back to over a decade. You might want to read this: http://en.m.wikipedia.org/wiki/Clariion
thereisnothinglikeabmw-gmail.com
10 Posts
0
June 3rd, 2015 02:00
Cool, that makes sense, thanks for your reply! I guess it arrays 5 years ago weren't as efficient
thereisnothinglikeabmw-gmail.com
10 Posts
0
June 3rd, 2015 03:00
Oh I know, but was referring to the AX4-5 I was working on.