Please help, EVA 3000 disk group problem

[Frode Olsen]

We have an EVA 3000, HSV100, version 3110.

We have 2 disks groups, 16x72GB, and 32x146GB. We want to add 8 more disks to the 146GB. However, when we try to add all 8, we get an error message:
"Insufficient Available Data Storage". So the only way I have been able to add more disks to the group is to add 2 of of the 8 disks. And after it has leveled out, I cannot add any of the remaining 6 disks to the group (same error message). We did realize we had 0% space left in the disk group prior, so after a lot of work we have created over 20% disk space left in the 146GB group (the 72GB group has 10% left). We currently have in the 146gb group (34 disks now) a capacity of 4373.3 GB, and occupancy is at 3437.67. The 146GB group only has 3 vdisks. I have searched this forum and docs, but cannot find anyone with a similar problem. Looking for some input. Thanks!

[Mark Poeschl_2]

That error is not one I'm familiar with either, but one thing you might try:

What is the disk failure protection level on your disk groups? You might try temporarily changing it to 'single' or even 'none', add the disks, and then returning the protection level to its current setting.

[Frode Olsen]

It is set to single. I'll try and set it to none and try again. Thanks.

[Sivakumar MJ._1]

As Mark Poeschl mentioned try with option 'none'.

If you have any snapshots configured then try deleting it, that should gain some space.

[Pieter 't Hart]

look at this thread : http://forums13.itrc.hp.com/service/forums/questionanswer.do?admit=109447627+1240387307840+28353475&threadId=1141800

This suggests with 146G disks there should be 297GB free space available.
"twice the amount of space of the largest disk in the diskgroup. In your case that would be 146gb * 2 = 292gb
Plus the 5gb extra 297gb."

Pieter

[Frode Olsen]

as I said I do have plenty of space available.

Now, quick question about setting the disk failure protection to None from Single. Might be a stupid question, but we want to make sure. Since this is vraid 5, and if we set it temporarily to None, we can still have 1 disk fail and we'd be ok? Can anyone clarify this? We have some very important servers on this disk group, so I assume setting the disk protection level to None, it'll need to level out again (took 24 hrs last time), so I don't want to risk losing a disk in this timeframe.

[IBaltay]

Hi, yes but with NONE, you immediately fall into the degraded state, so it is not a good idea to run the production with the NONE disk state. Try the single disk protection instead. It will reserve the size of 2 virtual spares for the possibility to copy the data from the failed disk to the virtual spare then ungroup the disk and thenafter you are not risking the to have another disk faiulre in the degraded state which would lead to the data loss...

[IBaltay]

errata:
Hi, yes but with NONE, after the disk failure, you immediately fall into the degraded state, so it is not a good idea to run the production with the NONE disk state. Try the single disk protection instead. It will reserve the size of 2 virtual spares for the possibility to copy the data from the failed disk to the virtual spare then ungroup the disk and thenafter you are not risking the to have another disk faiulre in the degraded state which would lead to the data loss...

[Frode Olsen]

It is already set to single disk protection, but we get the error when trying to add more disks. We have discussed this and we're considering leaving this alone, instead of trying to set it to None, as we do have important production servers using this disk group (ESX virtual servers). We have another EVA 3000 where we will have more space available in a couple of months.

[IBaltay]

it is true that internaly the physical disks in the DG are organized into the RSSs (redundant storage sets with 8 physical disks by default), so theoreticaly you can survive also the double disk failure if it occurs in 2 different RSSs, but certainly it is not good idea to risk it with the production data...