SNAPSHOT creation problem

sha · ‎2016-08-12

We're having a problem with one of our NetApp volumes: it's no longer creating any snapshots. It used to work.

SnapMirror for the volume is working properly.

Our Senior Storage Admin left us a few weeks ago, and my manager and I are absolute novices where DataOnTap is concerned. We haven't found IBM to be that helpful, so far, so we figured we'd give this forum a try.

This is the description of the problem my manager gave to IBM:

**************************************************************************

System type: Netapp FAS2200 NAS

Problem: PROCUREMENT Volume will not take daily/weekly snapshots, message saying "snapshot auto-delete or volume autosize still in progress"

attempted solution: The fs_size_fixed setting is currently set to "on" for this volume. It is part of a snapmirror relationship, and the snapmirror appears to be working. None of our other Snapmirror volumes have the fs_size_fixed setting set to "on", so I tried changing the fs_size_fixed setting to “off” for the PROCURMENT volume to try and get the snapshots working. So I enabled the setting, and had some weird behavior:

----------------------This is what the messages looked like before making the change.:----------------------------

NAS01> Wed Aug 3 17:28:46 EST [NAS01:wafl.snap.create.skip.reason:notice] : PROCUREMENT: skipping creation of hourly snapshot (snapshot auto-delete or vol ume autosize still in progress)

NAS01> Wed Aug 3 17:30:32 EST [NAS01:wafl.vol.autoSize.fail:info]: Unable to grow volume 'PROCUREMENT' to recover space: Volume has the fixed filesystem size option set.

----------- I then turned fs_fixed_size off. The volume started adding 10GB every 1 minute. ------------

Wed Aug 3 17:35:47 EST [NAS01:wafl.vol.autoSize.done:info]: Automatic increase size of volume 'PROCUREMENT' by 10485760 kbytes done.

Wed Aug 3 17:36:47 EST [NAS01:wafl.vol.autoSize.done:info]: Automatic increase size of volume 'PROCUREMENT' by 10485760 kbytes done.

Wed Aug 3 17:37:47 EST [NAS01:wafl.vol.autoSize.done:info]: Automatic increase size of volume 'PROCUREMENT' by 10485760 kbytes done. vol staus -b

Wed Aug 3 17:38:17 EST [CRPNAS01:wafl.vol.autoSize.done:info]: Automatic increase size of volume 'PROCUREMENT' by 10485760 kbytes done.

----- i then had to go increase the DR Procurement volume on DRNAS01 by the same size manually, because the replication started failing. -----------

---i then turned fs_fixed_size for NAS01 PROCUREMENT back to on-----------

So I fail to understand what is going on here, this behavior is the opposite of what I expected.

***********************************************************************************************************

Any suggestions you can give would be appreciated. We'll try to answer any question you have, but we may need to come back to you first and ask "How?"

Thanks in advance.

prakash_csc · ‎2016-08-13

Hi Sha,

This is a know issue, you have to check volume options in both source and destination.

Is there any other Snapmiror relationship having similar problem ?

raj_shrivastava11 · ‎2016-08-15

Hi Sha,

Why u have snapshot_autodelete and vol_autogrow on? Is it as per standard in your company?

is this a only volume you have or u have any other volumes where things are working fine?

Check volume options and try to disable snapshot_autodelete option first.

fs_size_fixed option should be off, then only u can resize your volume.

Let me know how it goes with above options

BR

Raj

sha · ‎2016-08-16

The PROCUREMENT volume is the only one having this problem.

Neither my manager nor I know exactly how / why the snapshot_autodelete and vol_autogrow options should be set; this was handled by our now-departed Storage Admin. Can you explain what they do and how they work, ie, how you think they ought to be set?

prakash_csc · ‎2016-08-16

Hi Sha,

Vol_autogrow option will come into effect when you disable fs_size_fixed to "off"

Volume size will be incremented automatically when source volume size increases.

But first, try to turn "off" snap autodelete .

If you have all these three options enabled , you would end up in issues like this.

sha · ‎2016-08-17

My manager wanted me to post this:

Shawn, can you post this sample on your message board post? The responses you’ve gotten indicate that we should turn snap_autodelete off , but from what I am seeing, it is already off. Maybe with this , the helpful posters can see what settings we need to change on the Procurement volume.

PROCUREMENT online raid_dp, flex nosnap=off, nosnapdir=off,

sis minra=off, no_atime_update=off,

64-bit nvfail=off,

ignore_inconsistent=off,

snapmirrored=off,

create_ucode=off,

convert_ucode=off,

maxdirsize=41943,

schedsnapname=ordinal,

fs_size_fixed=on,

guarantee=volume, svo_enable=off,

svo_checksum=off,

svo_allow_rman=off,

svo_reject_errors=off,

no_i2p=off,

fractional_reserve=100,

extent=off,

try_first=volume_grow,

read_realloc=off,

snapshot_clone_dependency=off,

dlog_hole_reserve=off,

nbu_archival_snap=off

Volume UUID: 8510b26a-952f-11e1-9f9f-00a0984d2b05

Containing aggregate: 'AGGRPR_01'

Plex /AGGRPR_01/plex0: online, normal, active

RAID group /AGGRPR_01/plex0/rg0: normal, block checksums

Snapshot autodelete settings for PROCUREMENT:

state=off

commitment=try

trigger=volume

target_free_space=20%

delete_order=oldest_first

defer_delete=user_created

prefix=(not specified)

destroy_list=none

Volume autosize settings:

state=on

maximum-size=6 TB

increment-size=10 GB

SALES online raid_dp, flex nosnap=off, nosnapdir=off,

sis minra=off, no_atime_update=off,

64-bit nvfail=off,

ignore_inconsistent=off,

snapmirrored=off,

create_ucode=on,

convert_ucode=off,

maxdirsize=41943,

schedsnapname=ordinal,

fs_size_fixed=off,

guarantee=volume, svo_enable=off,

svo_checksum=off,

svo_allow_rman=off,

svo_reject_errors=off,

no_i2p=off,

fractional_reserve=100,

extent=off,

try_first=volume_grow,

read_realloc=off,

snapshot_clone_dependency=off,

dlog_hole_reserve=off,

nbu_archival_snap=off

Volume UUID: 9ae624da-952f-11e1-9f9f-00a0984d2b05

Containing aggregate: 'AGGRPR_01'

Plex /AGGRPR_01/plex0: online, normal, active

RAID group /AGGRPR_01/plex0/rg0: normal, block checksums

Snapshot autodelete settings for SALES:

state=off

commitment=try

trigger=volume

target_free_space=20%

delete_order=oldest_first

defer_delete=user_created

prefix=(not specified)

destroy_list=none

Volume autosize settings:

state=on

maximum-size=1 TB

increment-size=1 GB

prakash_csc · ‎2016-08-17

Hi Sha,

All options to me seems good, just turn off fs_fixed_size option on PROCUREMENT volume. Do Not enable it back.

Try changing the snapshot schedule, match snapshot schedule with other volumes. Turn off fs_fixed_size option and snapshot schedule.

Increase your volume size by +50GB and leave it. See the response and let us know.

Thank you