MSA2060 optimal Configuration of Diskgroups

[ruahr]

Hello everyone,

I have a question about the MSA2060 optimal DiskGroup config for SMA-DP+ Raid.

HW Config: 60 HDD of 14TB (1 Controller-Shell an 4 Disk-Shelfs)

- Addon for 2 Disk-Shelfs in the future for growing capacity

We wll use a Dual Controller with 2 Pools.:

What it better per Pool:

1 DiskGroup with 30 Disks or 2 DiskGroups with 15 Disks.

If I understand it correct the 2 DiskGroups with 15 Disks  the netto ratio will be better than 1 DG with 30 Disks.

Thank you very much for your advise.

[venkat_y]

Hello@ruahr,

Yes, correct.
Two Disk Groups with 15 Disks each with MSA-DP+ is a good choice. If you create two disk groups with 15 disks each, you will have high fault tolerance and can easily expand your storage. HPE MSA 2060 storage can maximize performance and capacity by using two 15 disks with MSA-DP+ disk groups per pool.

To view more information on Raid  MSA-DP+ use the below link:
https://support.hpe.com/hpesc/public/docDisplay?docId=a00128580en_us&docLocale=en_US (Page no 22)

Hope this helps!
Regards,
Venkat

[support_s]

Hello,

Let us know if you were able to resolve the issue.

If you have no further query, and you are satisfied with the answer then kindly mark the topic as Solved so that it is helpful for all community members.

Please click on "Thumbs Up/Kudo" icon to give a "Kudo".

Thank you for being a HPE valuable community member.

[JonPaul · Accepted solution]

@ruahr @venkat_y 
I would argue differently from Venkat_y.  A single MSA-DP+ disk-group would be a better choice than 2x MSA-DP+ disk-groups.
First, a Pool can only contain 4x MSA-DP+ disk-groups
The MSA-DP+ disk-group is different than traditional RAID levels and allows for incremental expansion.
The MSA-DP+ disk-group has internal spare capacity split over all the disks, therefore rebuild times reduce as the spindle count increases
The MSA-DP+ disk-group has a high overhead for parity and spare.  The default spare capacity is the 2x the largest drive in the disk-group, it is recommended to increase spare capacity as the number of spindles increases.
Here are 2 documents on best practices:
  Best Practice Guide:  https://www.hpe.com/psnow/doc/a00105260enw
  Virtual Storage Technical Guide:  https://www.hpe.com/psnow/doc/a00103247enw
And you can also use the automated MSA Sizer tool to see what it would recommend:  https://ninjaonline.ext.hpe.com/msa-sizer/

[ruahr]

To all,

thanks for the replay. I created 2 DGs (1 DG for each Controller)

Regards

Ruediger

[JonPaul]

@ruahr 
Sounds good,  there are definitely some good reasons to create 2 disk-groups, 1 for each pool.
- Separation of workloads
- Possible increased performance due to the additional controller being used (unlikely in the configuration described as the spindle count will be the performance limiter)
- Logical/Physical separation of volumes
...
There are numerous reasons to configure differently which are hard to explore all in forum note.