MSA Storage
cancel
Showing results for 
Search instead for 
Did you mean: 

Re: MSA 2040 unmap done but still much more allocated in storage then in vsphere

 
SOLVED
Go to solution
Lieven_Embr
Visitor

MSA 2040 unmap done but still much more allocated in storage then in vsphere

I have a volume SAN02-vm04-BULO with 5.46 TB capacity,

Vsphere shows that 1.84 TB is in use and 3.62 TB is free. (65%)

i unmapped the free space. 

When I checked the file system I only find the VMs, nothing else.

When I doublecheck on the msa storage (volumes) however I see that 4036 GB is allocated.

What could be the cause of this and how do I deal with this?

 

# show disk-group-statistics
Name Time Since Reset Reads Writes Data Read Data Written Bps
IOPS I/O Resp Time Read Resp Time Write Resp Time
-------------------------------------------------------------------------------
dgA01 4294967 763346016 702591977 174.5TiB 66.5TiB 4667.5KiB
65 31667 31667 33018

Pages Allocated per Min Pages Deallocated per Min Pages Reclaimed
Pages Unmapped per Minute
------------------------------------------------------------------
0 0 0
0

rcA1 4294967 1086926489 339159506 74.7TiB 47.3TiB 43.9MiB
436 149 149 128

Pages Allocated per Min Pages Deallocated per Min Pages Reclaimed
Pages Unmapped per Minute
------------------------------------------------------------------
0 0 0
0
-------------------------------------------------------------------------------
Success: Command completed successfully. (2021-01-03 12:36:37)

 

4 REPLIES 4

Re: MSA 2040 unmap done but still much more allocated in storage then in vsphere

@Lieven_Embr 

What is the Controller Firmware version running in this MSA?

Check from Command Line how much showing as allocated at the Volume level.

show volumes details SAN02-vm04

I hope you have already done unmap command on datastore level from ESX CLI.

esxcli storage vmfs unmap -l SAN02-vm04-BULO

or

esxcli storage vmfs unmap -u <UUID_datastore>

Then keep checking below outputs from MSA end. Allocated size should reduced slowly. Please space reclaim will happen in 4MB pages so it's very slow as this is entry level array.

show volumes details SAN02-vm04

show disk-group-statistics

show pool-statistics

 

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
Lieven_Embr
Visitor

Re: MSA 2040 unmap done but still much more allocated in storage then in vsphere

Controller firmware is GL225P002-02 (2).

Unmap was indeed done.

# show volumes details SAN02-vm04
Pool Name Total Size Alloc Size Class
Serial Number WR Policy Cache Opt Read Ahead Type
Tier Affinity Snap Retention Priority Snap-Pool Desc
WWN Role Health Reason Action
-------------------------------------------------------------------------------
A SAN02-vm04 5587.9GiB 4033.7GiB Virtual
00c0ff28a74f00002dd65e5901000000 write-back standard Adaptive base
Performance medium
600C0FF00028A74F2DD65E5901000000 OK
-------------------------------------------------------------------------------
Success: Command completed successfully. (2021-01-04 15:41:23)

 

Shows 4 TB in use but in vmware there are no virtual machines or files in this volume.

 

# show disk-group-statistics
Management Controller is refreshing stale data. Please wait...
Waiting for system to finish computing health...
Name Time Since Reset Reads Writes Data Read Data Written Bps
IOPS I/O Resp Time Read Resp Time Write Resp Time
-------------------------------------------------------------------------------
dgA01 4294967 773321111 711763498 177.4TiB 67.9TiB 12.6MiB
103 27359 27359 42529

Pages Allocated per Min Pages Deallocated per Min Pages Reclaimed
Pages Unmapped per Minute
------------------------------------------------------------------
0 1 0
0

rcA1 4294967 1099115562 344673124 75.7TiB 47.7TiB 6301.5KiB
156 901 901 1124

Pages Allocated per Min Pages Deallocated per Min Pages Reclaimed
Pages Unmapped per Minute
------------------------------------------------------------------
0 0 0
0
-------------------------------------------------------------------------------
Success: Command completed successfully. (2021-01-04 15:43:45)

# show pool-statistics
Pool Pages Allocated per Min Pages Deallocated per Min
Pages Unmapped per Minute
----------------------------------------------------------
A 0 1
0

Time Since Reset Reads Writes Data Read Data Written Bps IOPS
I/O Resp Time Read Resp Time Write Resp Time
----------------------------------------------------------------------------
4294967 1872477420 1056459368 253.1TiB 115.6TiB 40.5MiB 426
14643 4270 41539

----------------------------------------------------------
Success: Command completed successfully. (2021-01-04 15:46:30)

 

 

As an experiment I uploaded the logs to MSA Storage Health Check

on san01 this gives:

- Enclosure 1 - Controllers GL225P002-02 (2)
- Enclosure 2 - IOMs 0147 (2) recommended 0149
- Drive Model - ST400FM0403 C006 (2) recommended 0007

on san02 this gives

:- Enclosure 1 - Controllers GL225P002-02 (2)
- Drive Model - ST400FM0403 C006 (2) recommended 0007
-  Scrubbing Enabled: AVAIL Drives, Global Spares, or Dedicated Spares found in array but Disk Background Scrub Not Enabled

 

Is it possible that the unmapping is still going on

Or should I delete the volume and recreate it?

Shawn_K
HPE Pro

Re: MSA 2040 unmap done but still much more allocated in storage then in vsphere

Hello,

In one of the MSA SMU screenshots you show there are two host groups mapped to the volume.

BACKUP_SERVERS and ESX_HOSTS.

I suggest you confirm that all the servers are not showing data on this volume, as it appears you have multiple hosts accessing and writing to the volume. I would especially check the back up servers as they are likely the ones adding to the volume.

I work for Hewlett Packard Enterprise. The comments in this post are my own and do not represent an official reply from HPE. No warranty or guarantees of any kind are expressed in my reply.

Cheers,
Shawn


I work for HPE

Accept or Kudo

Solution

Re: MSA 2040 unmap done but still much more allocated in storage then in vsphere

@Lieven_Embr 

My first observation, I can see SAN02-vm04 volume allocated size reduced from 4036.3GiB to 4033.7GiB which is 2.6GiB or in other way I can say approx 666 pages got reclaimed where each page size 4MB so reclamation going on at the Storage backend it seems. You can closely monitor show volume command output and see if the allocated size keeps decreasing or not.

My second observation, you have presented SAN02-vm04 to two servers or group of Hosts named BACKUP_SERVERS and ESX_HOSTS. I can see you have given read-write access to both. So you need to check on all those servers where this volume presented how you are utilizing this at the block level. Specially with BACKUP_SERVERS because you have not shown us any screenshot how volume usage looks from those servers Operating System level. We have seen only from ESX VMFS perspective.

 

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