EVA Disk Protection Level.

[Munees]

Hi Champs!

I have a disk group with 24 disks with single protection level in EVA. which means 2 out of 24 will be set as spare drives. My question is whether any 2 physical disks will be dedicated as spare drive or it virtually takes space from each disks?

Thanks in advance..

Cheers,
Munees

[Patrick Terlisten]

Hello,

the space is reserved over all disks in the disk group, there are no dedicated spare disks in an EVA. The space is reserved for rebuild, but as long as there is free space in the diskgroup, the EVA will use free space in the DG for rebuild.

Regards,
Patrick

[Sivaramakrishna_1]

Hi,

Actually 2 Disk means, not Disks. It will spare the 2*Max Size HDD space on the storage.

Suppose if u have a mixture of 146 and 300 and 450 in a single DG, it will reserve the 2*450GB Size on the DG.

[Munees]

Hi,

Thanks for your replies.

I got another qustion now. If the DG consists of mixture size of disks, how the space managed for disk protection? Is it eqally allocated from each disk or depends on disk size?

[Johan Guldmyr]

If you have a mixture of disk sizes in a disk group, for single disk protection it reserve 2x size of the largest disk.

[Víctor Cespón]

The "disk drive failure protection" option is only to set a soft barrier so you can be sure to have enough free space on the disk group to be able to rebuild all RAIDs.

The reconstructing process uses the free space on all the disks to rebuild the RAIDs. If you set the option to None and you have the space equivalent to 2 disks free, you're on the same situation as setting the paramenter as Single.

[Suhag]

I gone through the entire post and also found very usefull information. I want to post my scenario here as we are facing the problem releted to Disk protection level.

We are using EVA6400. We have created single disk group consisting of 17 no. of 1 TB disk . I have created some LUN/vDisk  with RAID1 and some with RAID5 and disk protection level was single.

Due to some issue,  2 no of disks from the same array were failed and all the vDisk/LUN were showing red cross and eventually it was a big disaster and all the data were lost permenentely.

To overcome this issue in future, i have created disk group once again and this time i set the  Disk protection level to double. What i was thinking that by doing double disk protection, i can sustain (or save data)  up to the failure of two disk.

Is my thinking is right? that by doing double disk protection, i am going to save data in case of 2 no of HDD fail simulteneously.

Please explain

Thanks in Advance,

Suhag Desai

TCS

[hvhari]

Double protection level will ensure enough space is available for re-construction of data and ensure redundancy upto two disks failure without needing a physical replacement. But the failures can not be simultaneous, as it is depnding on the parity. (when the failed disks are from different RSS Sets in eva, then you have the chance of not losing the data, even if both failed at same time.)

[Uwe Zessin]

The so-called "protection level"

     DOES NOT !!

protect any data. It is only a space reservation.

An RSS (redundant storage set)

     DOES NOT !!

protect any data. It is used to limit the size of a failure domain.

The only data protection is in the virtual disk's VRAID-level (1,5,6).

Create a virtual disk in VRAID-0 and no amount of "protection level" or "RSS" can prevent data loss from the failure of a single disk drive!!!

[sumitds074]

Hi,

Can any1 tell me wat is the meaning of protection level?
Some say it increases the number of drives count when we increase the protection level. But some say it only increases space in a disk
Can any1 pls tell wat is the functionality of this in a simpler way?
Thanks,
Sumit

[Torsten.]

Trying to keep it simple ...

With single or double you reserve space within a diskgroup.

If a disk fails (and has RAID1 data on it), the data will be moved from this bad disk and the "partner" disk to another pair of disks, hence 2 other disks are needed. So "single" means the space of 2 disks reserved on all remaining (!) disks within the diskgroup.

If a disk failes, it will be recovered into the remaining available space first, but you have the additional "protection level" space, even if the array is already at 100%.

This means the array has 100% space for data + reserved spaced for recovery (single or double level).


Hope you understand ...