Raid 5, failing disk in an array of six.

[CENTRINO]

We have a System I with the following disk configuration and with an
anomalous disk unit.

IBM will replace faling unit as soon as possible.

The system is running OK

Disk array is RAID 5 and they are all assigned to the ASP number 1

I have a question ¿ What would happen if another disk unit fails ?

Unit Type (M) Used
1 4327 70564 27,0 DPY NO PROT
2 4327 52923 27,0 DPY NO PROT
3 4327 52923 27,0 DPY ANOMALOUS
4 4327 52923 27,0 DPY NO PROT
5 4327 70564 27,0 DPY NO PROT
6 4327 52923 27,0 DPY NO PROT

Briefly, what is obtained by running WRKDSKSTS + F11
Thanks in advance.

--
.

[Jonathan Bailey]

For my system 8 disks in Raid 5 all of them say status ACTIVE.
If you loose another disk before replacing the faulty one and also re-building the protection, then you would loose the data from both failed disks, which would mean just about every object on the system. Since I expect the libraries are also lost the whole system is lost in reality.
I'm surprised your disks show different values for Size(M) since they look like the same type.
Apparently raid 6 can cope with 2 failed disks, but of course you loose 2 units of diskspace from the raid group.

HTH
Jonathan.

[CENTRINO]

Thanks, disk replaced, all going well
Just to confirm that we have this disk array Jonathan, surprisingly
there are two of them with different sizes, perhaps it has something to do
with microcode mirroriong or something like that cóuld not ask IBMrs:

Unit Type Size ASP Type Status
1 4327 70564 1 DPY ACTIVE
2 4327 52923 1 DPY ACTIVE
3 4327 52923 1 DPY ACTIVE
4 4327 52923 1 DPY ACTIVE
5 4327 70564 1 DPY ACTIVE
6 4327 52923 1 DPY ACTIVE




"Jonathan Bailey" <jonatha...@tesco.net> escribió en el mensaje
news:fdd670b5-fb31-4356...@googlegroups.com...

[jse...@yahoo.co.nz]

On Wednesday, April 9, 2014 1:01:37 AM UTC+12, CENTRINO wrote:
> Thanks, disk replaced, all going well
>
> Just to confirm that we have this disk array Jonathan, surprisingly
> there are two of them with different sizes, perhaps it has something to do
> with microcode mirroriong or something like that cóuld not ask IBMrs:
>
>
> Unit Type Size ASP Type Status
> 1 4327 70564 1 DPY ACTIVE
> 2 4327 52923 1 DPY ACTIVE
> 3 4327 52923 1 DPY ACTIVE
> 4 4327 52923 1 DPY ACTIVE
> 5 4327 70564 1 DPY ACTIVE
> 6 4327 52923 1 DPY ACTIVE
>

The difference in size isn't terribly odd. The RAID stripes are only on 4 of the disks, hence 2 of them are showing the full capacity.

[Jonathan Bailey]

My understanding of raid is clearly limited. I had the basic idea that all the checksum data was evenly spread across the disks. Thats how ours show up. I probably gave up reading about it after that point.

Jonathan.