MSA 2040 New Spare Drive Creation from existing Pool

[Lekhpal]

Dear All,

We've a MSA 2040, and one disk has failed.  At present replacing the failed disk is not an option(working on it with mgmt also). Two VGs in this MSA are RAID6 type, so no spares avaliable system wide.  We have two pools and one has 8TB free space. Is it possible to shrink this pool and then get one spare disk out of this pool. Then re-create the DG in pool with one less disk, and use that freed disk as system wide spare.

Please advise. Much appreciated.

BR

[support_s]

System recommended content:

1. HPE MSA 2040 SAN Storage - Configuration

2. HPE MSA 2040 SAN Storage - Spare Parts

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

Thank you for being a HPE valuable community member.

[Lekhpal]

Thank you so much.

But this doesn't anwer my question unfortunately. 

[ArunKKR]

Hi,

I believe this is virtual Pool configuration.

MSA 2040 supports both virtual Pool and linear vdisks.

There are no options to shrink a disk group.

If there are 2 disk groups in same Pool you could delete one of the disk group provided the Pool is not over committed and there is enough free space in first pool to accommodate data in the other Pool.  This is not a recommended procedure as disk group deletion will initiate a vdrain process which is quite slow . The best option for you in this scenario would be to take a data back up of the volumes in the MSA from host server end and delete and recreate the entire pool.  https://support.hpe.com/hpesc/public/docDisplay?docId=a00018177en_us&docLocale=en_US

Please share the output of below commands after deleting serial number information to better understand your configuration.   1. show pools 2. show disk-groups 3. show volumes

[Lekhpal]

Hello Arun.  

Thanks for the reply and the suggestions.  

Here are the outputs,  rest outputs will be sharing soon.

# show volume-groups
Group Name Serial Number Type Number of Members
----------------------------------------------------------------------
LNVR XXXXXXXXXXXXXXXXXXXXXXXXXX000000 Volume 2

Pool Name Total Size Alloc Size Class Type Health Reason Action
-----------------------------------------------------------------------
B ABCDE01 34.0TB 30.4TB Virtual base OK
B ABCDE03 38.9TB 38.9TB Virtual base OK

# show volumes
Pool Name Total Size Alloc Size Class Type Health Reason Action
-----------------------------------------------------------------------------
B ABCDE01 34.0TB 30.4TB Virtual base OK
B ABCDE03 38.9TB 38.9TB Virtual base OK
A VM03 54.9TB 54.9TB Virtual base OK
B VMO4 39.9TB 39.9TB Virtual base OK
A DATASTORE1 2199.9GB 1797.0GB Virtual base OK
A DATASTORE2 2199.9GB 1274.2GB Virtual base OK

# show vdisks
Name Size Free Own Pref RAID Class Disks Spr Chk Status Jobs Job% Serial Number Spin Down SD Delay Sec Fmt Health
Reason
Action
------------------------------------------------------------------------------------------------------------------------------------------------------------
POOL1 59.9TB 1879.2GB A A RAID6 Virtual 12 0 64k FTOL XXXXXXXXXXXXXXXXXXXXXXXXX00000000 Disabled 0 Mixed OK

POOL2 59.9TB 5255.9GB B B RAID6 Virtual 12 0 64k FTDN XXXXXXXXXXXXXXXXXXXXXXXXX00000000 Disabled 0 512e Degraded

One disk in the RAID-6 disk group failed. Reconstruction cannot start because there is no spare disk available of the proper type and size.
- Replace the disk with one of the same type (SAS SSD, enterprise SAS, or midline SAS) and the same or greater capacity. For continued optimum I/O performance, the replacement disk should have performance that is the same as or better than the one it is replacing.
- Configure the new disk as a spare so the system can start reconstructing the vdisk.
- To prevent this problem in the future, configure one or more additional disks as spare disks.

DGB01 59.9TB 5255.9GB B B RAID6 Virtual 12 0 64k FTOL XXXXXXXXXXXXXXXXXXXXXXXXXX00000000 Disabled 0 512e OK

Does this seem linear or virtual.  Can we use volumecopy to do some movement.  What is best way here. Please advise.

[ArunKKR]

Hi,

This is a virtual pool configuration.
Pool A has single disk group of size 59.9TB
Pool B has two disk groups of size 59.9TB each.
The free space in the disk groups is quite less.
Deleting a disk group in Pool B is not possible in this configuration.
Volume copy and migrating volume to different Pool or volume is not going to help.
I am afraid the only option would be to purchase and replace the disk.

[Lekhpal]

Hi Arun.
Thanks for the details. If I understand you , the Pool B is one which has the failed disk, and deleting one of the disk group in it is not possible. 
How to check the free space in each pool. I thought one of them had 10TB free. Can we delete some stuff to make free space. How can we calculate what amount to delete etc. Can we remove some stuff from the pool which has no failed disk, and then make some space.  We need to know how much to delete to make free space in pool, and how much is required.

Please advise. Thanks again for your assistance and insights.  BR.

[ArunKKR]

Hi,

The array will have 2 Pools - Pool A and Pool B.
There is a disk group with the name Pool2 in the array which seems to be creating the confusion.
show pools command will show the Pool capacity and free space. in it. 
Pool B is created using 2 disk groups - Pool2 (size 59.9TB and free space 5255.9GB) and DGB01 (size 59.9TB and free space 5255.9GB). 
Pool B total size would be around 119.8TB and free space of around 10.5TB. 
DGB01 is a RAID 6 disk group. 
The failed drive is in POOL2 disk group. 
Since one drive has failed the disk group is in FTDN (Fault tolerant with a down disk)state.
You could delete the underlying volumes in Pool to create free space in the Pool.
You could identify the volumes under each Pool from the output of show volumes command.
DGB01 should have enough free space to accomodate the data in Pool2 disk group before it could be deleted.
As mentioned earlier and as mentioned in the advisory shared, disk group deletion and vdrain is not a recommended procedure.

https://support.hpe.com/hpesc/public/docDisplay?docId=a00018177en_us&docLocale=en_US   MSA 2040 SMU reference guide: https://support.hpe.com/hpesc/public/docDisplay?docId=c04220794&docLocale=en_US

[Lekhpal]

Hi Arun.   Many thanks again.

Here is the output of the show pools.

# show pools
Name Serial Number Class Total Size Avail Snap Size OverCommit Disk Groups Volumes Low Thresh Mid Thresh High Thresh Sec Fmt
Health Reason Action
------------------------------------------------------------------------------------------------------------------------------------------------------------
A 00xxxxxxxxxxxxxxxxxxxxxxx000000 Virtual 59.9TB 1879.2GB 0B Enabled 1 3 25.00 % 50.00 % 99.64 % Mixed
OK
B 00xxxxxxxxxxxxxxxxxxxxxxx000000 Virtual 119.8TB 10.5TB 0B Enabled 2 3 25.00 % 50.00 % 99.82 % 512e
Degraded A virtual disk group is missing one or more disks. - Ensure that spare disks are available. Reconstruction should start automatically.
- When the reconstruction is complete, replace the failed disk(s). (Look for event 8 in the event log to determine which disk(s) failed.)
- Disk groups that cannot find compatible spares will automatically move data to fault-tolerant components.

-----------------------------------------------------------------------

Now, Pool B has free space of around 10.5TB. And, failed drive is in Pool2 DG. As you said, we can delete the volumes in pool to create space in pool. So, I check the show volumes:

# show volumes
Pool Name Total Size Alloc Size Class Type Health Reason Action
-----------------------------------------------------------------------------
B ISSEC01 34.0TB 30.4TB Virtual base OK
B ISSEC03 38.9TB 38.9TB Virtual base OK
A ISVM03 54.9TB 54.9TB Virtual base OK
B LNVRO4 39.9TB 39.9TB Virtual base OK
A MSADATASTORE1 2199.9GB 1797.0GB Virtual base OK
A MSADATASTORE2 2199.9GB 1274.2GB Virtual base OK

So, which volumes from above are potential deletion candidate to make DGB01 have enough free space. And, so that eventually Pool2 can be deleted.  Just hypothetically want to understand, which specific Pool2(DG) and its underlying volumes we can delete.

Thanks a lot again.

[ArunKKR]

Hi ,

The volumes in Pool B is spread across the disk groups Pool2  and DGB01 disk groups.

The size of each disk group is around 59.9TB. 

The volume size should be with in 59.9TB to achieve your objective.

Pool B has 3 volumes (ISSEC01,ISSEC03 and LNVRO4) with sizes of around 34.0TB , 38.9TB and 39.9TB.

You will need to delete any of these 2 volumes to ensure that volume size is below disk group capacity of 59.9TB.

[Lekhpal]

Hi Arun.  Thanks a ton.  

One last thing to ask now is for the disk replacement.  Once the new disk is available, we just need to remove the one failed and at its place insert the fresh new disk.  Right.  Is that the procedure.  Or, the procedure is something different.

Best

[Lekhpal]

Hi.  And, also once the deletion is done,   whatever data exists in this VDG will drain out to bigger VDG.  After that what will be the steps.  How to create the new VDG with one spare there in.   Thanks again for excellent inputs.  Best.

[ArunKKR · Accepted solution]

Hi,

I guess the drive replacement query is with the current status ie, all disk groups in current state,

Once the drive has been replaced it needs to be configured as global spare so that it kicks in as part of the disk group in degraded state and rebuild starts.
If Dynamic Spare Configuration option is enabled the drive would kick in automatically as soon as drive is replaced.

Below command could be used to verify current Dynamic Spare Configuration status:

show advanced-settings

Once the disk group deletion has been completed all the disks that were part of the disk group would appear in available state.
These available disks could be used to create a new disk group under Pool A or Pool B with the number of disks of your choice.

You might refer to the MSA SMU guide for detailed instructions and understand MSA storage architecture:

https://support.hpe.com/hpesc/public/docDisplay?docId=c04220794&docLocale=en_US

[Lekhpal]

Hi Arun.  show disks "usage" is showing virtual for all the disks and one is showing as Failed..Dynamic sparing is disabled.

So it needs to be configured as global spare so that it kicks in.  Got it.  And, other part is also now clear.  Best