MSA P2050 warning disk space....

[WillyEB70]

Dear all,
I have some questions about the configuration and disk space usage of an MSA P2050 (Fibre Channel) system.

This system initially had 8 x 2TB drives. The elements in the system were A disk group in radi 6, a pool A and one volume Vol0001. There were no Global Spare Disks.

I recently added 4 more 2TB disks, created a disk group in raid 6 and then extended the volume Vol0001 by about 4000GB.

The current configuration is as follows:

# show volumes
Pool Name     Total Size Alloc Size Type Large Virtual Extents  Health Reason Action
-------------------------------------------------------------------------------------
A    Vol0001  15.3TB     6660.7GB   base Disabled               OK
-------------------------------------------------------------------------------------
Success: Command completed successfully. (2022-11-14 15:40:22)
# show pools
Name Serial Number                    Blocksize Total Size Avail Snap Size OverCommit  Disk Groups Volumes  Low Thresh  Mid Thresh  High Thresh  Sec Fmt   Health     Reason Action
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
A    ***removed*** 512       15.9TB     9.3TB 0B        Enabled     2           1        50.00 %     75.00 %     98.65 %      512n      OK
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Success: Command completed successfully. (2022-11-14 15:40:24)
# show disk-groups
Name   Size     Free     Pool Tier    % of Pool  Own RAID   Disks Status Current Job Job%      Sec Fmt   Health     Reason Action
----------------------------------------------------------------------------------------------------------------------------------
dgA01  11.9TB   8657.0GB A    Archive 75         A   RAID6  8     FTOL                       512n      OK
dgA02  3994.3GB 663.9GB  A    Archive 24         A   RAID6  4     FTOL   VRSC      87%       512n      OK
----------------------------------------------------------------------------------------------------------------------------------
Success: Command completed successfully. (2022-11-14 15:40:27)
# show disks
Location   Serial Number         Vendor   Rev   Description Usage         Jobs Speed (kr/min)  Size     Sec Fmt    Disk Group Pool Tier    Health
------------------------------------------------------------------------------------------------------------------------------------------------------
1.1        ***removed***   SEAGATE  N003  SAS MDL     VIRTUAL POOL       7               2000.3GB 512n       dgA01      A    Archive OK
1.2        ***removed***   SEAGATE  N003  SAS MDL     VIRTUAL POOL       7               2000.3GB 512n       dgA01      A    Archive OK
1.3        ***removed***   SEAGATE  N003  SAS MDL     VIRTUAL POOL       7               2000.3GB 512n       dgA01      A    Archive OK
1.4        ***removed***   SEAGATE  N003  SAS MDL     VIRTUAL POOL       7               2000.3GB 512n       dgA01      A    Archive OK
1.5        ***removed***   SEAGATE  N003  SAS MDL     VIRTUAL POOL       7               2000.3GB 512n       dgA01      A    Archive OK
1.6        ***removed***   SEAGATE  N003  SAS MDL     VIRTUAL POOL       7               2000.3GB 512n       dgA01      A    Archive OK
1.7        ***removed***   HGST     CD05  SAS MDL     VIRTUAL POOL       7               2000.3GB 512n       dgA01      A    Archive OK
1.8        ***removed***   HGST     CD05  SAS MDL     VIRTUAL POOL       7               2000.3GB 512n       dgA01      A    Archive OK
1.9        ***removed***   HGST     C907  SAS MDL     VIRTUAL POOL  VRSC 7               2000.3GB 512n       dgA02      A    Archive OK
1.10       ***removed***   HGST     C907  SAS MDL     VIRTUAL POOL  VRSC 7               2000.3GB 512n       dgA02      A    Archive OK
1.11       ***removed***   HGST     C907  SAS MDL     VIRTUAL POOL  VRSC 7               2000.3GB 512n       dgA02      A    Archive OK
1.12       ***removed***   HGST     C907  SAS MDL     VIRTUAL POOL  VRSC 7               2000.3GB 512n       dgA02      A    Archive OK
------------------------------------------------------------------------------------------------------------------------------------------------------
Info:  * Rates may vary. This is normal behavior. (2022-11-14 15:40:31)

The virtualization system, oVirt, that uses this storage now meets a LVM volume of roughly 14TB used like this:

[root@infm-vm04 log]# pvdisplay
  --- Physical volume ---
  PV Name               /dev/mapper/3600c0ff0003c465b6510f85b01000000
  VG Name               ebe2403c-a0c0-4bde-9691-25e86119b068
  PV Size               13.99 TiB / not usable 94.00 MiB
  Allocatable           yes
  PE Size               128.00 MiB
  Total PE              114643
  Free PE               17185
  Allocated PE          97458
  PV UUID               teBzaf-MwYo-sy0c-IsD3-XIXR-3dlB-IQnfWx

After making these changes,, however I found this message in the P2050 events:

2022-11-11 11:05:12 [473]  #A330:  MSA 2050 SAN Array SN#00C0FF3ACB50 Controller A INFORMATIONAL A volume is using more than its threshold percentage of its virtual pool. (pool: A, SN: ***removed***) (volume: Vol0001, SN: ***removed***) (used percentage: 41.67%, threshold percentage: 10%) (pages used: 1587944, total pages: 3810361)

Moderator edit: Removed the Serial numbers for privacy. 

Can you explain why P2050 displays 41.67%  "used percentage" ? 

Have you got any idea ?

Thanks a lot.

Best regards

Willy

[JonPaul]

@WillyEB70 
An unfortunate known issue.  When you create a volume it certain instances it will set the 'threshold-percent-of-pool' value to 10%.  There is no way to change this after the fact.  You can see this value in the API output of the CLI. 
CLI>  set cli-parameters api
CLI> show volumes Vol0001
...
<PROPERTY name="threshold-percent-of-pool" type="string" size="16" draw="false" sort="size" display-name="Threshold Percent of Pool">10.00 %</PROPERTY>
...

It does not affect the Pool thresholds which will warn you of the pool filling. The INFORMATIONAL event can be ignored.  Your Volume is using ~6.6TB of 15.TB which is the ~40% of the pools

Please note that your configuration does not conform to best practices.  The 2 different disk-groups will have different performance characteristics and could result in performance fluctuations at your application.  See MSA Best Practice guide and MSA Virtual Storage Technical Reference.

[WillyEB70]

I know that two sets of different records aren't the best thing but at the moment I couldn't do otherwise.

I would like to understand why P2050 shows 41.67%.. 

Thanks again

Bye

[JonPaul · Accepted solution]

@WillyEB70 
Sorry I buried that reasoning in the middle of the other stuff.

The entire volume is 15.3TB  the  allocated usage is 6660.7GB (6.6TB)
6.6TB / 15.3TB == 0.43 == 43%  there's a bunch of rounding issues in that calc so +/- 2% is pretty good
If you grab the pages used from the event message you will see:
1587944 pages / 3810361 pages == 0.4167 == 41.67%

Now if the question is why does it show so much space used eventhough I haven't moved that much data....
Is your filesystem doing a zero fill of the volume?

if the host writes ZEROs to the MSA, the MSA will store those zero blocks.  It will then over time re-claim the zero blocks and show smaller used capacity.
You can see this using the statistics information:   
CLI>  show pool-statistics
CLI>  show tier-statistics tier all
Both of these pages deallocated -- reclaimed -- if data is being deleted and the TRIM command is being sent or if ZERO pages are being reclaimed you will see pages being deallocated.