IBM Storage The online community where IBM Storage users meet, share, discuss, and learn. Join the Community
When hot spare disk takes over in distributed raid 5 or 6 configuration?
Your suggestions will be highly appreciated.
There is a very handy Blog post on this here: https://barrywhytestorage.blog/2020/01/09/some-details-about-how-draid-works/
The short answer is that Distributed RAID stripes the data in packs across the pool of all drives in a manner such that every drive has data and spare block capacity. This means that when a drive fails, every drive in the array contains data, parity and spare capacity and will participate in the read, regenerate and write of the data from the lost disk.
So unlike traditional raid where only a single drive is used for the write , you see better general performance as all drives are part of the repair and you aren't limited to the write speed of a single spare drive during rebuild. I hope that helps
I guess I don't understand the question. In DRAID5/6 there is no single drive that is a "hot spare" there is spare space on every drive of the DRAID array (as well as data, P and Q). When a drive fails, all drives participate in the rebuild and accept data from the failed drive in their "spare" areas. When the failed drive is replaced, data is built back on the new drive. That drive will contain data, parity, Q and spare space.