vFiler migration with flex-clones fails - Volume is a clone and scanners have not completed on the parent volume: cannot execute operation

[fletch2007]

It takes almost 6 hours to run the initial snap mirror sync for this vFiler migration, and its failed twice with

Waiting for SnapMirror-Resync to complete on 'dbdev2, dbdev5, dbdev7' flex-clones…

then

SnapMirror Resync failed: Reason: SnapMirror error: Volume is a clone and scanners have not completed on the parent volume: cannot execute operation

Src and dest are 3270 clusters running 8.1RC2

This is the last remaining migration blocking our long awaited shelf upgrade - and we have a case open with Netapp

Anyone seen this one and worked around it?

Thanks,

Fletcher

[adaikkap]

What version of dfm are you using ?

Regards

adai

[fletch2007]

Hi Adai - here are the versions of DFM and NMC we are using

thanks

Application name: NetApp Management Console

Version: 3.1

Build number: 3.1.0.4506

Installation directory: C:\Program Files (x86)\NetApp\Management Console

Java VM OS Architecture: x86

OS Name: Windows Vista

OS Version: 6.1

Java VM Name: Java HotSpot(TM) Client VM

Java Version: 1.6.0_13

Java Vendor: Sun Microsystems Inc.

Java Free Memory: 44.9 MB

Java Total Memory: 207.6 MB

Java Max Memory: 508.1 MB

DFM Version: 5.0.0.7636 (5.0)

DFM Serial Number: 1-50-008475

DFM Host Name: dfm-05

DFM Node Limit: 250 (currently managing 5)

Provisioning Node Limit: 250 (currently managing 5)

Protection Node Limit: 250 (currently managing 5)

DFM OS: Microsoft Windows 2003 Service Pack 2 (Build 3790) x86 based

DFM CPU Count: 2

DFM System Memory: 4095

DFM Install Directory: C:/Program Files/Network Appliance/DataFabric/DFM

DFM System Time: Feb 20, 2012 07:52:27 AM

[tanmoy]

Looks like some background scanners are still running on the clone volumes, e.g. block reclamation scanner which typically gets kicked off after snampirror baselining and in some other conditions. Another process is deswizzling which takes place after snapmirror initialization. I am doubting one these things is happening here. If I remember correctly there is some way to reduce the scanner frequencies. We need to find out the exact reason, first.

[fletch2007]

I found a vestigal snapshot/snapmirror relationship for one of the volumes being migrated and cleared it

We'll see in a few hours when the baseline completes if this made a difference

thanks

[fletch2007]

Failed again - same error:

na02: SnapMirror Resync failed: Reason: SnapMirror error: Volume is a clone and scanners have not completed on the parent volume: cannot execute operation

What is the workaround for this?

thanks

[fletch2007]

Migrating another vFiler with flexclones - this time under 8.1 and latest Core and NMC 3.0.1

The baselines are done, but NMC is stuck "in progress" - job status reports warnings:

Waiting for the deswizzling to complete on the parent volume 'flexcloneXYZ' of  to be completed

Will this eventually succeed or hang indefinately?

thanks

[scottgelb]

I didn't know that it waited for deswizzling to finish (target volume on the mirror).  It will finish but depending on how many snaps that can take a while.. you can see deswizzling with 'priv set advanced ; wafl scan status'

[fletch2007]

thanks Scott -

parent and 3 flexclone status below - can I suspend snaps to help things out?

Volume stride:

Scan id                   Type of scan     progress

   61342    active bitmap rearrangement     fbn 31043 of 36319 w/ max_chain_len 3

Volume stridedbdev3:

Scan id                   Type of scan     progress

  108832    active bitmap rearrangement     fbn 34394 of 36059 w/ max_chain_len 3

Volume stridedbdev4:

Scan id                   Type of scan     progress

  108837    active bitmap rearrangement     fbn 34778 of 36059 w/ max_chain_len 3

Volume stridedbdev5:

Scan id                   Type of scan     progress

  108842    active bitmap rearrangement     fbn 34029 of 36059 w/ max_chain_len 3

[fletch2007]

I opened another case with support - seeing these

SnapMirror resync of stridedbdev3 to 3270head1:stridedbdev3 : Volume is a clone and scanners have not completed on the parent volume:    cannot execute operation.

and the swizzling looks like it its looping - will suspending snaps help?

Volume stride:

Scan id                   Type of scan     progress

   61342    active bitmap rearrangement     fbn 32693 of 36319 w/ max_chain_len 3

Volume stridedbdev3:

Scan id                   Type of scan     progress

  108832    active bitmap rearrangement     fbn 36044 of 36059 w/ max_chain_len 3

Volume stridedbdev4:

Scan id                   Type of scan     progress

  108837    active bitmap rearrangement     fbn 368 of 36059 w/ max_chain_len 3

Volume stridedbdev5:

Scan id                   Type of scan     progress

  108842    active bitmap rearrangement     fbn 35679 of 36059 w/ max_chain_len 3

[scottgelb]

Deswizzling can be sped up deleting source snaps so less to deswizzle on the target. I would get support input too in case of a Burt or other workaround.

Sent from my iPhone 4S

[KUBER_KATANA1]

Hey,

So what was the solution for this?

I am facing the same issue .. can't resync and swizzling running for the 2 clones.

thanks in advance.

Kuber