Disk Protection level Vs Raid 5 Redundancy level

[Navaneeda]

I have a EVA 4000 and configured with a single disk group. I have given Disk Group protection level set to 1. So 2 Disks space are going to use for the single protection level.
I have Virtual Disks Configured in Raid Level 5.

Now my question is, in the above setup if i have a single disk failure then there is no data loss and disk group protection will take care of the disk failure. In this setup normally Raid 5 will work with single disk failure... so where my raid 5 redundancy is used in the above setup ???

[IBaltay · Accepted solution]

Hi,

The difference is in the disk safety/availability enforcement:

in the disk failure the data are taken over to the spare drive space (failed disk ungroup from the DG to the spare space) to avoid the double disk failure in the RSS (because in case of the 2 disk failure in one RSS (both in RAID1 and RAID5 the whole DG becomes inoperative.
Then after the failed disk replacement, the new disk is groupped to the DG and then the data from the spare disk space are taken back to it....

If you dont have any spares (disk protection set to NONE), in case of the disk failure you are running in the RAID5 degraded mode and there is not any mechanism to reduce the risk of the second disk failure...

[IBaltay]

ERRATA - the sentence
"there is not any mechanism to reduce the risk of the second disk failure..."

should be
"there is not any mechanism to reduce the risk of the data loss in case of the second disk failure..."

[Navaneeda]

Thanks for the Reply,

From the reply i understand the below.
If a Single or two disk failure will be taken care by DG Protection. If my DG protection Set to None and I have Raid 5 Vdisk, in this case if one Disk Failure in my Raid 5 Vdisk will survive without any data Loss.If more than one disk failure in my Raid 5 will loose the array config and data Loss. Pls correct me if im wrong.

Can you pls explain me about the RSS in the Disk Group.I know 8 Disks forms and create one RSS Set. How the RSS is working ????

Thanks in Advance

[IBaltay]

Hi,
1.
yes you are right,
when you have the disk protection other then none (single or double), the degraded state of the DG is very short because of the spare drive take over and ungroup algorithm of the EVA controller...

2.
this thread goes deeper in RSS (it is for the EVA5k, but the RSS algoritm was not changed):
http://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=886230