snap manager for hyper-v failed on vm backup?

[#1]

Hi,

Snapmanager for hyper-v inewly s installed with a snapinfo path cluster disk non CSV Netapp LUN

When trying to create a backup with Snapmanager it fails with the following:-

SnapManager for Hyper-V backup failed to complete

Backup of the Dataset Name: test

Backup id: 6b2eb748-2656-4128-9cf7-8dd8fae37c36 failed to execute

Error :An error occured while invoking SnapDrive for processing snapshot rename.

Object reference not set to an instance of an object.

Backup status of the virtual machines involved in this backup operation are given below

Backup Failed for

VM Name: Windows2003-Test VM ID: 9B739109-9225-4A47-8DE8-4E41D4480EC9

----------------------------------------------------------------

Failed to rename the Snapshot copy ({D26411D2-5177-4C13-9CC5-E2E55BE3CE8C}) of the LUN to the new Snapshot copy name (test_HyperVCluster1_VMHOST1_06-02-2010_11.37.09).

LUN Name =

Storage Path = /vol/

Protocol Type =

Storage System Name =

Error code : The LUN may not be connected, because its mount point cannot be found.

-----------------------------------------------------

Any help would be appreciated

Kev

[#2]

Hi

Take a look at this KB article

https://now.netapp.com/Knowledgebase/solutionarea.asp?id=kb57876

Regards

Amrita

[#3]

Hi,

Thanks for the reply, it looks like you may be right, however converting the volumes to GPT has somehow screwed up the licensing.

Snapmanager for hyper-v seems to have switched to server based licensing!

Still working on it.

Thanks

Rgds

Kev

[#4]

Thanks for the prompt reply, I have now  changed the partition from mbr to gpt but now get a licensing issue!

Even though the filer has the required license and smhv has been reinstalled pointing to the filer for licenses

See below:-

Thanks

Rgds

Kev

[6/2/20106:05:03PM,(11)]Constructingbackuptasks [6/2/20106:05:03PM,(11)]Constructionofbackuptaskscompletedsuccessfully. [6/2/20106:05:03PM,(11)]Executingbackuptasks.. [6/2/20106:05:03PM,(11)]StartingexecutionofbackuptaskBackupTask[Host:vmhost2]
Request:Objectstobebackedup: Windows2003Test1[Hyper-V]
[6/2/20106:05:03PM,(11)]WaitingforexecutionofbackuptaskBackupTask[Host:vmhost2]
Request:Objectstobebackedup: Windows2003Test1[Hyper-V] tocomplete [6/2/20106:05:04PM,(18)]SnapManagerBackupReport
Objectstobebackedup: Windows2003Test1[Hyper-V]
[6/2/20106:05:04PM,(18)]Validatingbackuprequest [6/2/20106:05:04PM,(18)]Validationofbackuprequestcompleted [6/2/20106:05:04PM,(18)]ValidatingSnapManagerforHyper-Vbackuprequest... [6/2/20106:05:04PM,(18)]ValidationSnapManagerforHyper-Vbackuprequestcompletedsuccessfully [6/2/20106:05:04PM,(18)]Processingbackupofhyper-VVMs... [6/2/20106:05:04PM,(18)]Checkingif[HyperVCluster1]is a clusteredsystem.. [6/2/20106:05:04PM,(18)][HyperVCluster1]is a clusteredsystem.. [6/2/20106:05:04PM,(18)]Checkingiftherequestisre-directedfromownernodeofthecluster.. [6/2/20106:05:04PM,(18)]Therequestisre-directedfromownernodeofthecluster.. [6/2/20106:05:20PM,(18)]FailthebackupifVMcannotbebackeduponline?:False [6/2/20106:05:20PM,(18)]InitializingPolicyEngine [6/2/20106:05:20PM,(18)]====InitializingPolicyhandler:[Scriptsettings]==================== [6/2/20106:05:20PM,(18)]InitializingScriptexecutionhandler.. [6/2/20106:05:20PM,(18)]InitializationofPolicyhandler:[Scriptsettings]-Done
[6/2/20106:05:20PM,(18)]InitializationofPolicyEngine-Done [6/2/20106:05:20PM,(18)]Processingpolicies..
[6/2/20106:05:20PM,(18)]====Processingpolicy:[Scriptsettings]====================
[6/2/20106:05:20PM,(18)]Noscriptneedtobeexecuted. [6/2/20106:05:20PM,(18)]Processingofpolicy:[Scriptsettings]-Done
[6/2/20106:05:20PM,(18)]PreparingforbackupofWindows2003Test1[Hyper-V] [6/2/20106:05:20PM,(18)]RetrievingVSSComponentinformation... [6/2/20106:05:20PM,(18)]VSSComponentPath : \43E1EC0A-B1CB-4DF8-B620-0D8F7CC72DB1 [6/2/20106:05:20PM,(18)]RetrievalofVSSComponentinformationcompletedsuccessfully. [6/2/20106:05:20PM,(18)]Retrievingsnapinfoinformation... [6/2/20106:05:20PM,(18)]Retrievalofsnapinfoinformationcompletedsuccessfully. [6/2/20106:05:20PM,(18)]SnapManagerforHyper-VisnotlicensedonthehostorintheStorageSystem,backupisaborted [6/2/20106:05:20PM,(18)]SnapManagerforHyper-VisnotlicensedonthehostorintheStorageSystem,backupisaborted [6/2/20106:05:20PM,(18)]Backuptask:failedtocomplete [6/2/20106:05:20PM,(18)]AutoSupportisdisabled,AutoSupportupdatewillnotbeperformed [6/2/20106:05:20PM,(18)]Anerroroccuredwhilecreatingthehyper-Vbackup. [6/2/20106:05:20PM,(18)]Backuptaskfailedtosucceed.SnapManagerforHyper-Visnotlicensedonthehostorinthe StorageSystem,backupisaborted [6/2/20106:05:20PM,(18)]Anerroroccurredwhileprocessingthebackup. SnapManagerforHyper-VisnotlicensedonthehostorintheStorageSystem,backupisaborted
[6/2/20106:05:20PM,(18)]SnapManagerforHyper-VisnotlicensedonthehostorintheStorageSystem,backupisaborted [6/2/20106:05:21PM,(18)]<?xmlversion="1.0"encoding="utf-16"?> <VMStatusResultxmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <VMName="Windows2003Test1"Id="43E1EC0A-B1CB-4DF8-B620-0D8F7CC72DB1"Result="Failed"/> </VMStatusResult> [6/2/20106:05:21PM,(11)]CompletedbackupofsharedVMs

license+error.xps.zip ‏167 KB

[#5]

Hi

Could you please use SDw for creating LUN's? That is our recommendation to use SDW 6.2 for all LUN creation. And then could you please try out backup?

Regards

Amrita

[#6]

Hi Amrita,

I am in the same scenario.

The 2 node HyperV Cluster along with the VM's is already in place and I am doing a SMHV config.

I did create the shared snapinfo and dedicated report LUNS for each host using SnapDrive 6.2, but the CSV with VM's already existed.

SMHV keeps throwing these vague errors about licensing!

When I  move all VM's to a single host and create a dataset of all VM's , then the backup is successful from each host. ( Which means that both hosts see the SMHV filer  license )

However, when the VM's are owned by 2 nodes and are in a single Dataset, then the backup is successful on the node that I execute the backup from and fails on the other with license error. Is it by design that the SMHV backup is redirected to each host that owns the VM's even tough the all VM's belong to a single CSV?

Also, when I try to create a schedule, it fails in the last step when it tries to create the schedule for the other host with a credential error.

I tried SnapDrive and SMHV with local admin users and also with Domain users.

Not sure what the issue is. I have 2 support cases open for 3 weeks but to no avail.

Any help is highly appreciated.

Thanks

Animesh

animesh_pillai@hotmail.com

[#7 · Accepted solution]

H

Also I am not sure if you saw this in our release notes

You need to install SnapDrive for Windows on all hosts before installing SnapManager for Hyper-V.

If the hosts are members of a cluster, all nodes in the cluster require the installation of SnapDrive for

Windows.

When SnapManager for Hyper-V starts it communicates with SnapDrive for Windows to get the list

of all virtual machines running on a host. If SnapDrive for Windows is not installed on the host, this

API fails and the SnapManager for Hyper-V internal cache does not update with the virtual machine

information.

You might receive the following message:

Error :SnapManager for Hyper-V is not licensed on the host or in the Storage

System, backup is aborted:

The credentials issue can be because communication is not happening properly across the cluster so please check the cluster config too. Also in a cluster environment you should be using domain user and not local user.

Regards

Amrita

[#8]

Hi

In your case the workaround is to install SDW before SMHV. See my other post with the snippet from the release notes.

Regards

Amrita

[#9]

HI Amrita,

Thanks for the prompt response.

As part of the steps provided in the open support case , i uninstalled SMHV, Snapdrive and installed SnapDrive and then SMHV.

However, still get these errors.

The SDW service account is individual host account and not a domain/cluster account as of now. The same is the case with SMHV account.

However, in the past , I tried SDW and SMHV with domain admin account.

Thanks

Animesh

[#10]

Hi

So in the SMHV configuration did you add the cluster name? If the cluster name is added you need to give the domainname\administrator ( for all CSV. Cluster configurations). For a dedicated host a local account should be fine. I think you should also check the cluster configuration once to make sure that communication between the nodes is happening properly. Since a case is opened with NGS they will take a look at the complete configuration and let you know where things are failing exactly.

Regards

Amrita