Quantcast
Channel: Data Protection Manager - Hyper-V and CSV Clusters forum
Viewing all 628 articles
Browse latest View live
↧

DPM backup replica instead of original VM

October 12, 2014, 10:38 pm
$
0
0

Hello,

I have this situation:

1. Several standalone (no cluster) production Hyper-V server with VM's on it. Lets call them original VM's. 

2. Backup server, with Hyper-V role on it. One VM on this server is DPM. And on this server is target for VM replication of original VM's (replica VMs)

DPM is configured with VMM for support for live migration. 

I configure DPM to protect original VM. But if there is replication of VM on backup server, backup job fails with error that there is no DPM agent on backup server. If I install DPM agent on backup server - it back up not the original VM, but replicated VM.

One workaround is to uninstall VMM agent on backup server, but i need real solution.

How to tell DPM that it should backup original VM, but not the replica ?

↧
Search

Hardware Providers with DPM 2012 R2 and Hyper-V 2012 R2?

October 14, 2014, 9:38 am
$
0
0
With the changes to how DPM 2012 R2 and Hyper-V 2012 R2 makes an "online" backup, are hardware VSS providers necessary anymore? Or is there a benefit to disabling them on the 2012 R2 hosts? We are using an Equallogic PS6500x if that helps.
↧
↧

Hyper-V Cluster Hosts BMR Backups take over 12 hours

October 14, 2014, 2:15 pm
$
0
0

Hello, 

We are using 2012 R2 DC to create a Hyper-V  2 node cluster.  The backup is DPM 2012 R2 running on 2012 R2 DC.  The DPM system was upgraded to 2012 R2 from DPM 2010 about a year ago and has worked for the most part.

Two problems we are having.  I am using the VSS writer from the SAN manufacture.  On our old cluster I used the Microsoft VSS writer and had no problems unless I loaded the C: up with too many files on the cluster node.  I saw another post that says to disable the HW VSS writer but that was unsuccessful the last time I tried it.  I am going to try it again though.

1.  I currently get a random VM that hangs during the backup.  The VM still runs but I cannot migrate it and backups fail from that point on.  In Hyper-V backup the Status field shows "Backing up...".  I am waiting for this to happen on a non critical VM but alas that is not happening. <sigh>  To correct the problem I have to pull power from the cluster node the system was on and then about 90% of the time I have to replace to different .sys files in c:\windows\system23\drivers and then run the startup repair utility from the windows install disk.

2.  I did not notice this until this week but my BMR backups of the Hyper-V hosts are taking over 12 hours to complete.  The BMR on a different stand alone server is taking less than 10 min.  What is causing this and if these are taking this long is it interfering with the backup of the VMs on the cluster and causing the problem in paragraph 1 above?  I have tried to stagger my backups so they don't interfere with each other but if this one is taking 12-14 hours that will never happen.

↧

"DPM was unable to access the cluster resource group" while restoring a vm

August 20, 2014, 12:59 am
$
0
0

Hi there,

We're having the following Setup:

  • 3 Hyper-V 2012 R2 (core) Server clustered
  • 2 SOFS 2012 R2 (core) Server clustered
  • Data Protection Manager Server 2012 R2 on a physical box

We have no trouble backing up the VMs over the hyper-v Cluster but restoring the VM terminates with the following error (Event ID 3111 Source DPM-EM).

The recovery jobs for Online\Test-win81 that started at ----- with the destination of ----- have completed. Most or all jobs failed to recover the requested data. (ID: 3111)

DPM was unable to access the cluster resource group ---- for ----of protection group -. Recover the datasource using the alternate location workflow. (ID: 32579)

We've been so far unable to identify the cause. We can access the Cluster via Failover Cluster Manager from the DPM Server without a Problem.


Best Regards

System Engineer

↧

System Center 2012 R2 DPM - Recover the datasource using the alternate location workflow. (ID: 32579)

October 27, 2013, 10:05 am
$
0
0

Unable to recover a protected virtual machine to a Hyper-V 2012 R2 cluster using the System Center 2012 R2 DPM "Restore the VM to the original host or cluster" option.  The recovery fails with the following error message:

"The recovery jobs for Offline\VMName that started at Sunday, October 27, 1:01:07 PM, with the destination of SCVMM VMName Resources.ClusterName.Domain.Net, have completed.  Most or all jobs failed to recover the rquested data (ID: 3111)

DPM was unable to access the cluster resource group SCVMM VMName Resources.ClusterName.Domain.Net for Offline\VMName of protection group Protection Group Name.  Recover the datasource using the alternate location workflow. (ID: 32579)"

Recovery succeeds using the "Restore the VM to a different host or cluster", but this process requires additional post recovery cleanup and is less preferred to the "Restore the VM to the original host or cluster".

↧
↧

Snapshots in Hyper-V 2012r2 can't be deleted

October 16, 2014, 1:15 am
$
0
0

Hi,

we have a big problem with Snapshots, which are done by DPM 2012r2 over the Hyper-V Manager.

Our System Center Data Protection Manager backs up our whole system. Some Virtual Machines only via a Cluster Network Snapshot over the Hyper-V Manager.

Now we have the problem, that some VMs still have Checkpoints/Snapshots from the DPM-Backup and we could not delete it. But it'snecessary to delete it! If we click right on the Snapshopt/Checkpoint the options that are shown: Settings, Export, Rename and Help. There is no chance to click delete or remove!

In the eventlog is no information about that. In the moment it is very bad to reboot the Hosts or restart some cluster services! The VMs have to run and there is no time for downtime now.

Is it an error of the Hyper-v Manager? Or anyone a solution for this problem?

I hope that somebody could help me!

best regards

Huffelpaff

↧

Parallel Jobs

October 16, 2014, 2:30 am
$
0
0

Hi,

I have a bit of problem with DPM as the data is growing it is taking longer and longer for recovery point and tape backup.

I changed Max Parallel Job to 3 to run and finish quicker but it does not run parallel in my 2 node cluster. When I change it back to 1. it does throw an error "

DPM could not find a valid recovery point on disk. " There is no way I could set sequence on it. It is too moody to stick with the sequence as DataSourceGroups.xml file. Is there a way to make it better and faster . Where am I going wrong for parallel jobs?

I have already gone through few sites including this.

https://social.technet.microsoft.com/Forums/en-US/c40fa9f4-0162-410b-bb4b-ac4e3c576746/clarification-please-on-what-dpm-could-not-find-a-valid-recovery-point-means?forum=dpmtapebackuprecovery

↧

October 2014 update rollup for Windows Server 2012 R2 and Data Protection Manager 2012 R2

October 17, 2014, 9:17 am
$
0
0

Hello all,

After installing the October 2014 update rollup for Windows Server 2012 R2 (KB2995388) on our Windows 2012 R2 Core Edition backup jobs of System State start to Fail, below the generic error:

The replica of Non VSS Datasource Writer on hostname.domain is inconsistent with the protected data source. All protection activities for this data source will fail until the replica is synchronized with consistency check.

The job was cancelled. The user either cancelled the job or modified the associated protection group.

Removing the update rollup all jobs complete without errors. Somebody is expiriencing the same issue? How can we resolve the problem?

Thnks, Andrea

↧
Search

DPM 2012 R2 UR3 HV backup behaivor

October 19, 2014, 11:35 pm
$
0
0

Hi,

I have a dedicated DPM 2012 R2 UR3 to backup VMs located in HV 2012 Cluster using Software VSS.

I have added a Backup NIC card on each HV host and connect them to a dedicated Switch where the DPM server is also connected to isolate backup traffic from the production.

Each HV node have a MGMT, SAN, HBeat and Backup Network.

I started to test backup one VM to see the DPM behavior and I noticed the following:-

Backup NIC on the HV nodes have no traffic. DPM triggered the agent to start VM backup. VM hosted nodes started to move data through the Backup NIC card. The strange thing is that the Hbeat NIC was transferring the exact same amount of data that the Backup NIC card was transferring. I checked the Backup NIC card and it was used by DPMRA which is perfect. When I checked the Hbeat NIC card it was used by “System”. The normal Hbeat traffic before starting the VM backup was 800 Kbps and the Backup NIC card was 0. When the VM backup started both the Backup and Hbeat NICs started to transfer about 200 Mbps each. The reason that im worried is there is a chance that the Hbeat NIC traffic might be used intensively during the backup which cause the Hbeat to fail to Contact the cluster causing the HV node to restart and you know the rest.

Questions:-

  1. Why the DPM is using the Hbeat?
  2. What is the data path when DPM agent trigger the backup of the VM? DPM Server – HV Node – Storage.
  3. Where does the VM snapshot saved? And how it works.

Just have several questions more but answering these will start a discussion J

Thanks in advance,

Laith

Please remember to click “Mark as Answer” on the post that helps you. This posting is provided "AS IS" with no warranties. knowledge is valid only if it is shared by All.


↧
↧

Windows Server 2012 VM's can only be backed up using saved state

October 20, 2014, 11:45 am
$
0
0

Have a 2008R2 Virtual Host with 2012R2 VM's.  I'm trying to get DPM2012sp1 to backup these vm's using the "Backup Using Child Partition Snaptshot".  I've read at the link below that I need to create a registry key for this to be possible. I've done this, but I still do not have the option.  Any help would be appreciated.  

https://social.technet.microsoft.com/Forums/en-US/2ceb4087-d495-4406-ab0e-08c3405a764e/backup-of-windows-2012-child-hyperv-server-as-saved-state-instead-of-snapshot?forum=winserverhyperv

HKLM\Software\Microsoft\Windows NT\CurrentVersion\SystemRestore 
REG_DWORD:ScopeSnapshots:0x0


↧

DPM 2012 R2 Scheduled Jobs Not Running

October 22, 2014, 9:20 am
$
0
0

I' am trying to use DPM 2012 R2 to back up a Hyper-V cluster.  The cluster nodes are running Server 2012 R2.  When I initially created the protection groups, all the VM's backed up without incident.  I noticed later that there were no recovery points being created.  I could have DPM do a consistency check and then there'd be a recovery point.  When I watch the scheduled jobs in DPM, the time for a job will arrive, the job will be removed from the "Scheduled" list and nothing at all happens.  All the protection groups show the beautiful green ball with the checkmark in it and a big ol' "OK" next to it.  None of the scheduled jobs are even attempting to run on any of the VM's that this system is supposed to be backing up.  All the nodes in the cluster have all the latest MS updates and so does the DPM server.

Someone please help!!

Thanks

↧

SPM 2012 R2 RU3 doesn't see any VMs ( CSV Storage)

October 22, 2014, 1:10 pm
$
0
0

Hello Folks 

Here is the scenario in short :
DPM2012 R2 RU3 
Hyper-v Cluster on CSV storage ( windows 2012 R2 server - last update was June 2014  ) 
DPM agent is installed and status is OK 
When I add a protection Group for Hyper-v ; system enumerates ( takes a minute) and then shows only typical server 
backup options such as Volumes , shares and system state . there is only one extra entry for Hyper-v which is Hyper-V host. 
There is absolutely NO SIGN of any VMs !! I also checked to ensure that actually this host owns a CSV LUN and it does , it doesn't host Quroum disk . Hyper-v hosts 21 VMs right now but I don't see them in DPM and no errors . 

This is a supported setup :
http://technet.microsoft.com/en-us/library/hh758027.aspx

BTW ; during enumeration ; I get a message that seems to be informational :

DPM is unable to enumerate contents in \\?\Volume{88ef8033-dad1-11e3-80b5-806e6f6e6963}\ ,\\?\Volume{88ef8039-dad1-11e3-80b5-806e6f6e6963}\ on the protected computer DPHVHOST01.company.com. Recycle Bin, System Volume Information folder, non-NTFS volumes, DFS links, CDs, Quorum Disk (for cluster) and other removable media cannot be protected. (ID: 38)

Any thoughts ?

↧

VM replica creation stopping co-located VMs

October 24, 2014, 12:39 pm
$
0
0

I've created a protection group for a VM that is hosted on HyperV 2012 R2, the host has multiple VMs, including VMs that are co-located on the same LUN. My short term protection is to DISK. I'm running DPM 2012 R2.

When I create the new protection group and it begins to create the VM replica to disk (over the network, same subnet) all the other VMs on that host become unavailable, they show in the MMC as up but they stop responding to all attempts to connect, either RDP or HTTP (some are web servers). I've had the same situation happen when DPM was running a consistency check on an existing protected VM.

Is DPM supposed to lockup the host and its VMs when performing this kind of backup? Until recently I only backed up to tape, this is my first try at VM replicas. Thanks for any help.




↧
↧

DPM 2012 R2 Hyper-V Online vs Offline

October 27, 2014, 8:42 am
$
0
0

Hello,

I recently upgraded DPM from 2012 to 2012 R2.  And our Hyper-V servers were upgraded to 2012 R2 as well.

What affects whether a VM shows Offline or Online in DPM?

The obvious answer is whether it is on or not, but I have several VMs that are always on, and yet they show up as "Offline".

Are they getting backed up correctly if that is happening?   (Seems like the OS of the ones that are showing as "offline" is either 2008 or 2008 R2.)

Thanks.

↧

Incomplete restore from inside VHDX File

October 31, 2014, 12:13 pm
$
0
0

I am running DPM2012R2 inside a Hyperv-2012R2 VM.

I am trying to restore a directory from a 2008R2 VM running on the same Hyper-V Server.  When I try to restore a directory from inside the vhd, the restore is incomplete even though I see all the data I need from inside the DPM console.  Dedupe must be on somewhere because when I check the folder properties in the 2008r2 server, it totals to 2.5TB but the drive in explorer shows a fixed 1TB drive and the VHD is also 1TB.  I am not running dedupe on the volume that stores my VHD files so I am not sure where dedupe is running.  Is there any way to restore the files that are displayed in the DPM console? 

Fred Pinto


↧
Search

Replica is inconsistent

November 3, 2014, 6:52 am
$
0
0

I have System Center 2012 R2 installed on Windows server 2008. I am trying to protect a VM that is running SQL Server 2008. The VM is being managed by a Failover Cluster in Windows 2012 R2 standard and has 4 VHDx's attached which are located on a SAN. Now, when ever I try to run an consistency check on that protection group member, I receive the following error:

________________________________________________

The replica of Microsoft Hyper-V \Online\VM on "Target.com" is inconsistent with the protected data source. All protection activities for data source will fail until the replica is synchronized with consistency check. You can recover data from existing recovery points, but new recovery points cannot be created until the replica is consistent.

For SharePoint farm, recovery points will continue getting created with the databases that are consistent. To backup inconsistent databases, run a consistency check on the farm. (ID 3106)

The VSS application writer or the VSS provider is in a bad state. Either it was already in a bad state or it entered a bad state during the current operation. (ID 30111 Details: VssError:The writer experienced a non-transient error.  If the backup process is retried,
the error is likely to reoccur.
 (0x800423F4))

________________________________________________

Now when I check the event log from the server in question I receive 5 separate event logs pointing to Hyper-v-VMMS.

The First:

'VM' cannot create the storage required for the checkpoint using disk C:\TARGET\TARGET\VM.vhdx: General access denied error (0x80070005). (Virtual machine ID 97333BCA-130E-4F20-934A-CB26A19DDBF6) - Event ID 16370

The Second:

Checkpoint operation for 'VM' failed. (Virtual machine ID 97333BCA-130E-4F20-934A-CB26A19DDBF6) - Event ID 18012

The Third:

Could not create backup checkpoint for virtual machine 'VM': General access denied error (0x80070005). (Virtual machine ID 97333BCA-130E-4F20-934A-CB26A19DDBF6) - Event ID 10150

The Fourth:

VSS writers inside virtual machine 'VM' failed to perform BackupComplete to its shadow copy (VSS snapshot) set: Catastrophic failure (0x8000FFFF). (Virtual machine ID 97333BCA-130E-4F20-934A-CB26A19DDBF6) - Event ID 10172

The Fifth:

The operation failed. - Event ID 16010

________________________________________________

As suggested online, I ran vssadmin list writers from an elevated command prompt and every writer appears to be stable with no errors. The weird thing is, I am not receiving any errors from the other 3 VHDX's. In the recovery tab, there are no history of any recovery point for that VM. I am able to protect the VM databases but not the VM itself. I also gave full rights in the folder where all the VHDx's are located. Any ideas?

Thank you
↧

VM in locked State

November 4, 2014, 10:12 am
↧
↧

Unable to select a Hyper-V guest that was previously a Hyper-V 2012 Replica

November 5, 2014, 2:36 am
$
0
0

We have a Hyper-V 2012 environment which uses Hyper-V replication and I have configured DPM 2012 to back up the Primary servers only (I know it is neither supported nor required to backup the replicas of the VMs).

This has worked very well until recently when we had to start the replica of one of the servers due to a hardware failure of the Hyper-V host server.  The switch over to the replica was successful however I was unable to back it up in DPM as a Hyper-V guest as it was not visible when expanding the HyperV section of the server when trying to select it (only the Host Component is available).

I put this down to the fact that the server I was trying to select was still the replica but now the original host has been repaired and the original 'primary' VM has been reconfigured as the new 'replica' and the original 'replica' configured as the new 'primary'.

Despite this 'role reversal' of the primary and replica VMs I am still unable to see the new 'primary' VM in the 'Available members' section when modifying the protection group.  We have created a test server on the same host to see if it was something to do with the agent, but that does appear so it looks as though the agent is ok (I can select other resources from the server anyway).

Any help or advice relating to this would be much appreciated,

Thanks

Chris

↧

Noisy SCOM alerts when Offline backup a Hyper-V VM

November 6, 2014, 1:42 am
$
0
0

Dear Forum Members,

I would ask you the following question, maybe there are some experienced users who have already seen this problem :/ The issue is that we back up a lot of VMs on Hyper-V via DPM Offline backup, which will bring down the VM to saved state for an amount of time. Unfortunately this (VM is not accessible) triggers some SCOM alerts which are of course not needed, they are just noise. Do you have any ideas how we could get rid of these unneccessary alerts? Or for example automaticaly put these VMs into maintenance mode during the back up process?

Thank you really for your help!

Best Regards,

Chris

↧

Backup of Hyper-V guest with more than one virtual disks failes with SCDPM 2012R2

November 11, 2014, 6:48 am
$
0
0

Hello,

We have moved guests from 2008 R2 Hyper-V cluster to the 2012R2 Hyper-V cluster. All guests were backed up with SCDPM2010 perfectly. Now we try to backup with SCDPM 2012R2 and some guests backup fails constantly so we can't take even one backup. After short investigation we have found that all "problematic" guests have two virtual disks without any exception. On host we get error events like:

Log Name:      Microsoft-Windows-Hyper-V-Worker-Admin
Source:        Microsoft-Windows-Hyper-V-Worker
Date:          11/10/2014 1:33:13 PM
Event ID:      3280
Task Category: None
Level:         Error
Keywords:     
User:          NT VIRTUAL MACHINE\3CD65890-39BE-41A3-A2A7-B7752166BBBB
Computer:      aaa.com
Description:
'BBB111' could not initiate a checkpoint operation: Element not found. (0x80070490). (Virtual machine ID 3CD65890-39BE-41A3-A2A7-B7752166BBBB)

Log Name:      Microsoft-Windows-Hyper-V-VMMS-Admin
Source:        Microsoft-Windows-Hyper-V-VMMS
Date:          11/10/2014 1:33:13 PM
Event ID:      18014
Task Category: None
Level:         Error
Keywords:     
User:          SYSTEM
Computer:     aaa.com
Description:
Checkpoint operation for 'BBB111' was cancelled. (Virtual machine ID 3CD65890-39BE-41A3-A2A7-B7752166BBBB)

So the question is if somebody faced this issue and what is workaround.

Thank you,

Victor

 

↧
Viewing all 628 articles
Browse latest View live
Search
<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>