HPE Storage Users Group
https://3parug.net/

What causes the high snapshot space consumption ?
https://3parug.net/viewtopic.php?f=18&t=2809		 Page 1 of 1
Author:  georg.pohl [ Wed Feb 28, 2018 8:19 am ]
Post subject:  What causes the high snapshot space consumption ?
Hi Community,

here is my second post in this user group.
I do not get clear about the numbers i see with space utilization.

HPE 3PAR OS 3.2.2 MU4 with peer persistance .
We have one TPVV that holds a Test VM.
This VM is existing vor about 9 month now.
Its soley task is to see how RMC-V works.
No productive action on this volume or VM.
In this 3PAR volume there is a VMFS5 datastore only containing the above vm.

The Recoveryset is set to hourly snapshot, and hold them for 7 days, summing up to 167 recent snapshots, we see.
(I wonder where the 168th is, but thats a different story… 24*7=168)


The copy cpg is put to a separate cpg called CP_CPG_R6_FC_NL_MA

Now, since beginning of february i receive emails like this:
Component: Virtual Volume 464 RMC_Test CPG 3 CP_CPG_R6_FC_NL_MA Domain 0 - Short Dsc: VV RMC_Test snapshot space warning Event String: Reserved snapshot space for VV RMC_Test has reached allocation warning of 160G (80% of 200G)

I was wondering why i receive this message as no action is on that vm,
therefore no blocks should be freezed as changes inside virtual copies and
therefore, i should have no consumption or at least only a few bytes, even not GB.

XXX-3PAR-MA cli% showcpg CP_CPG_R6_FC_NL_MA
---------------(MB)----------------
----Volumes---- -Usage- -- Usr --- ---- Snp ---- -- Adm ---
Id Name Warn% VVs TPVVs TDVVs Usr Snp Total Used Total Used Total Used
3 CP_CPG_R6_FC_NL_MA - 28 28 0 0 28 0 0 433920 406016 8192 0
---------------------------------------------------------------------------------------
1 total 0 28 0 0 433920 406016 8192 0

XXX-3PAR-MA cli% showcpg -d CP_CPG_R6_FC_NL_MA
---------------(MB)----------------
----Volumes---- -Usage- -- Usr --- ---- Snp ---- -- Adm --- --- LD ---- - RC_Usage -
Id Name Warn% VVs TPVVs TDVVs Usr Snp Total Used Total Used Total Used Usr Snp Adm Usr Snp
3 CP_CPG_R6_FC_NL_MA - 28 28 0 0 28 0 0 433920 402816 8192 0 0 2 2 0 0
----------------------------------------------------------------------------------------------------------------
1 total 0 28 0 0 433920 402816 8192 0 0 2 2 0 0

XXX-3PAR-MA cli% showvv RMC_Test
----Rsvd(MB)----- -(MB)-
Id Name Prov Type CopyOf BsId Rd -Detailed_State- Adm Snp Usr VSize
464 RMC_Test tpvv base --- 464 RW normal 1152 239104 24960 204800
-------------------------------------------------------------------------------
1 total 1152 239104 24960 204800

XXX-3PAR-MA cli% showvv -s RMC_Test
---Adm--- ------------Snp------------ ----------Usr-----------
--(MB)--- ----(MB)----- --(% VSize)-- ---(MB)---- -(% VSize)-- -----(MB)------ -Capacity Efficiency-
Id Name Prov Type Rsvd Used Rsvd Used Used Wrn Lim Rsvd Used Used Wrn Lim Tot_Rsvd VSize Compaction Dedup
464 RMC_Test tpvv base 1152 839 235904 223645 109.2 80 0 24960 19808 9.7 80 0 262016 204800 147.5 --
---------------------------------------------------------------------------------------------------------------------------
1 total 1152 839 235904 223645 24960 19808 262016 204800 147.5 --

vSphere reports 114 GB free of 200 GB on datastore view.
Windows inside VM shows 15 GB occupied.

The CP_CPG is on nearline.
XXX-3PAR-MA cli% showsys -space -devtype NL shows
…
Snapshot : 1365682
Used : 763996 --> 746 GB
Used (Bulk VVs) : 0
Unused : 601686
…

All of the 64 (32 x base violume, 32 x .r volume) volumes we have are distributed across three different copycpgs.
All 64 volumes are configured to use copycpgs in these mentioned three provisioned copy cpgs
The other two copycpgs show about 50 GB of space utilization but do not even have snaps.

My interpretation is:
a) The volume can grow up to 204.800 MB alias 200 GB
b) 19.808 MB alias 19,3 GB are in use by the base volume (out of 24.960 MB reserved).
c) if necessary snapshots can occupy 235.904 MB for this volume;
in reality snapshots use already 223.645 MB alias 218,4 GB and therefore we have an overprovisiong of about 9.2 % related to the base volume size.

Still, I do not understand, wherefrom these huge comsumption comes from, as the vm does nothing but to be present (perhaps some windows updates…)

d) why then does the CPG use 402.816 MB alias 393,375 GB.

There are no other VMs or Datastores that are touched by any kind of RMC* operation.

I do not get the numbers together.
Certainly, i do oversee soemthing, but what ?
Can anybody shed some light ?
Author:  sjm [ Wed Feb 28, 2018 9:53 pm ]
Post subject:  Re: What causes the high snapshot space consumption ?
you need to compact the cpg used for snapshots on a regular basis to return that unused space back to that CPG.I See no mention of this
Page 1 of 1 All times are UTC - 5 hours
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/