HPE 3PAR StoreServ Storage
cancel
Showing results for 
Search instead for 
Did you mean: 

HPE 3PAR 8450-Sparing Policy and frame size

 
SOLVED
Go to solution
Highlighted
Occasional Visitor

HPE 3PAR 8450-Sparing Policy and frame size

Hello All,
I have HPe3PAR 8450 storage frame with 6 * 1.92TB SSD. Please confirm is sparing policy impact the data capacity of frame.


if i choose minimal it took 1787 spare chunklets. (one disk as a spare)
and if i choose Default it double the chunklets 3574 (two disk as a spare)

But when I try to create CPG in both sparing policy configuration I am getting same capacity for CPG 3985 GB.

Q.1 Why this is happening, if two disk use as a spare in Default sparing policy, my CPG capacity should be decreased by 1787 chunklet size.

Q.2 With 6 disk and minimal spare policy, is it recommended to create CPG with Raid5 and set size 5D + 1P? What HPe best recommendation

5 REPLIES 5
Highlighted
HPE Pro

Re: Sparing Policy and frame size

HI GUPSAN

Thank you for coming on HPE Comunity forums.

It depends on a lot of factors as the the data provided here is very minimal . I  would recommend to runs the default and not minimal . Also HPE recommends Raid6. He does not have many drives. We would also need to know how many are present in the array and how is it distributed across.

I would suggest you to kindly get in touch with HPE Technical support and log a case and get this query resolved.

Hope this helps !!

Thanks
Mahesh202

I work for HPE

Accept or Kudo


[Any personal opinions expressed are mine, and not official statements on behalf of Hewlett Packard Enterprise]
Highlighted

Re: Sparing Policy and frame size

@GUPSAN 

As per my understanding your total Capacity of the System (6 x 1787) = 10722 chunklets or 10722GB

Out of 10722 chunklets, if you select Spare Algorithm as Default then your raw free space of the system will be (10722 - 3574) = 7148 chunklets and 3574 chunklets will go as reserve spare chunklets

If you select Spare Algorithm as Minimal then your raw free space of the system will be (10722 - 1787) = 8935 chunklets and 1787 chunklets will go as reserve spare chunklets

When you create any CPG then you are actually considering space 7148 chunklets or 8935 chunklets as your data space. All calculation will happen based on CPG parameter defined.

You can check below outputs if you want to understand more about space,

showpd -c

showsys -space

You can refer below best practices whitepaper and concept guide to get more details,

https://psnow.ext.hpe.com/doc/4aa4-4524enw

https://support.hpe.com/hpesc/public/docDisplay?docId=c04204225

 

Hope this helps!
Regards
Subhajit

I am an HPE employee

If you feel this was helpful please click the KUDOS! thumb below!

***********************************************************************


I work for HPE
Accept or Kudo
Highlighted
Occasional Visitor

Re: Sparing Policy and frame size

Thanks  for answering my first Doubt. 

Will you please also  provide answer of my 2nd Q.

 

With 6 disk and minimal spare policy, is it recommended to create CPG with Raid5 and set size 5D + 1P? What HPe best recommendation

 

Why I am asking this ?

 

If using Raid5(5+1) spread in a frame having only 6 disks

That means that you need 6 chunklets to build a stripe => 6 disks are used as it is forbidden for a stripe to have 2 chunklets sharing the SAME disk !

 

So : What happen in case of a drive fail => where to recreate the missing 6th chunklet ? as only 5 disks surviving, 2 chunklet would use the same disk !

You see what I mean ?

 

Thanks.

Highlighted

Re: Sparing Policy and frame size

@GUPSAN 

There is something called HA cage and HA mag where HA means High Availability.

At the time of CPG creation you need to select HA option as Cage or Mag or Port. 

Roughly I can tell you that when you select HA cage, in this case at the time of LD creation chunklet getting selected from each drive and all drives will be located in different cage. So in case of entire cage failure you data will be safe.

You can go through Concepts guide RAID section and you will get more details.

 

Hope this helps!
Regards
Subhajit

I am an HPE employee

If you feel this was helpful please click the KUDOS! thumb below!

***********************************************************************


I work for HPE
Accept or Kudo
Highlighted
Solution

Re: Sparing Policy and frame size

Hi GUPSAN,

You are right. If you have six disks and a CPG that uses a six-disk set size (ssz), if a physical disk fails, then logical disk changes will hang. As will application writes.

To handle physical disk failures, you want the CPG set size to be less than the number of physical disks.

And HPE Best Practices strongly recommend using RAID 6.

Options:

  1. (recommended) Add a pair of 1.92TB SSD, use RAID 6 with ssz 6 (4+2). Tunesys to balance existing usage across new drives before creating or extending volumes.
  2. (not recommended, but still an option) With the existing six disks, use RAID 5   8^(   with ssz 5 (4+1) and be very careful when replacing a drive that a wrong (good) one is not pulled by mistake.
    Which is why HPE Best Practices strongly recommend RAID 6.

Note: While I am an HPE Employee, all of my comments (whether noted or not), are my own and are not any official representation of the company

Accept or Kudo