Data Backup and Recovery

LUN Space Reservation: disabled, but not giving space back to the volume

txskibum2000
9,017 Views

So, I have a weird issue...

I have a 135Gb Volume called "ctxlic_db".  The volume guarantee is enabled, with 0 fractional_reserve.

I have a 100Gb LUN called "ctxlic_db.lun" in the volume with space reservation disabled.

LUN is attached to a Win2003 host.  The windows host shows the drive/lun is using just under 7Gb (93Gb free).

The 135Gb Volume, with 43Gb of Snapshots, show 88% full, using 118Gb space!?

So, my question is, why is my volume showing 88% full, when LUN Space Reservation is disabled, and the host is using only 7Gb.

I would generally expect the volume used space being 50Gb (7Gb used space + 43Gb of snapshots).

Details below and attached.

Thanks!

fa02> vol status /vol/ctxlic_db

         Volume State           Status            Options

      ctxlic_db online          raid_dp, flex     nosnap=on, create_ucode=on,

                                                  convert_ucode=on, no_i2p=on,

                                                  fractional_reserve=0

                Containing aggregate: 'fc_aggr1'

fa02>

fa02>

fa02> lun show -v /vol/ctxlic_db/ctxlic_db.lun

        /vol/ctxlic_db/ctxlic_db.lun  100.0g (107389255680)  (r/w, online, mapped)

                Comment: "I"

                Serial#: W-OPKZYMnCUu

                Share: none

                Space Reservation: disabled

                Multiprotocol Type: windows_gpt

                Maps: viaRPC.iqn.1991-05.com.microsoft:ctxlic.aaefcu.org=5

fa02>

fa02>

fa02> df /vol/ctxlic_db/ctxlic_db.lun

Filesystem              kbytes       used      avail capacity  Mounted on

/vol/ctxlic_db/      141557760  124558928   16998832      88%  /vol/ctxlic_db/

/vol/ctxlic_db/.snapshot          0   45575696          0     ---%  /vol/ctxlic_db/.snapshot

fa02>

fa02>

fa02>

fa02> df -h /vol/ctxlic_db/ctxlic_db.lun

Filesystem               total       used      avail capacity  Mounted on

/vol/ctxlic_db/          135GB      118GB       16GB      88%  /vol/ctxlic_db/

/vol/ctxlic_db/.snapshot        0GB       43GB        0GB     ---%  /vol/ctxlic_db/.snapshot

fa02>

fa02>

1 ACCEPTED SOLUTION

gridley
9,017 Views

No worries, happy to (try to ) help!

As I recall there are things that will stop SR from running. such as snapshots being taken. I will have a search and see if anything useful turns up.

It might be worth opening a case if you have time, so we can get you your storage back!

View solution in original post

8 REPLIES 8

gridley
9,017 Views

Have you written data to the LUN and then deleted it? You will need to run space reclamation from SnapDrive if so to get the space back.

txskibum2000
9,017 Views

This is what I was thinking, thank you for your reply.

I tried to run Space Reclamation twice (via Snapdrive), and it would get to 3-4% (10-15minutes), then quit running.

gridley
9,018 Views

No worries, happy to (try to ) help!

As I recall there are things that will stop SR from running. such as snapshots being taken. I will have a search and see if anything useful turns up.

It might be worth opening a case if you have time, so we can get you your storage back!

txskibum2000
9,017 Views

Thanks.

I have read the "Reasons for SnapDrive to automatically stop Space Reclaimer", and none of those reasons apply to me.

I will open a case with NetApp.

gridley
9,017 Views

EDIT: just saw your reply, so this isn't relevant!

From the docs:

https://now.netapp.com/NOW/knowledge/docs/snapdrive/relsnap63/html/software/admin/GUID-1F3DD258-F601-4CAE-B765-05507F76A493.html

"SnapDrive for Windows automatically stops Space Reclaimer in several instances.

  • During any LUN management operation on a LUN running Space Reclaimer, including LUN disconnect and LUN delete operations
  • During any Snapshot copy management operation except Snapshot copy rename and delete
  • On all LUNs of the same storage system volume during Snapshot copy creation
  • If the SnapDrive service is stopped
  • During LUN restore operations for any volume mount points directly or indirectly mounted (cascading) from the disk being restored
  • During Windows host cluster failoverIf a host cluster failover operation occurs on a Windows Server 2003 MSCS or Windows Server 2008 failover cluster while Space Reclaimer is running on a shared LUN, space reclamation will stop running on that LUN.
  • During any MPIO path management operations, including adding or removing an initiator or active path selection"

Did any of this happen?

jakob_bena
9,017 Views

hi,

you can use the option of your snapdrive tool called spacereclaimend. this will refresh the used space of your volumen.

on filer site you can use wafl reallocate, but i think for you it is the best way to do this with the snapdrive option.

regards

txskibum2000
9,017 Views

Support has refered me to KB 3011577, which i have read, and has suggested to create a new LUN...  NDMP copy the data from the old LUN to the new LUN, and attached to the new LUN -destroy the old LUN.  I figure this would be the recommendation...

Thanks all!

gridley
9,017 Views

Completely random thought - I wonder if the LUN move functionality would achieve the same resullt? Probably something to test in a different environment... I imagine not since it uses mirroring instead of an NDMP copy.

Public