Got questions about SMVI?

[mlisa]

Our team is very interested in the questions posted on SMVI functionality, configuration, and automation … thanks for taking the time to share.  And please keep the questions and feedback on your experiences working with SMVI, good or bad, coming.  Good feedback makes for a good day, bad feedback makes for a better product J.

I also wanted to let you know that NetApp is hosting an SMVI/SRM Webcast on February 19 focusing on data protection in a VMware environment: NetApp SMVI for backup/restore and VMware SRM for disaster recovery. SDDPC Sys Admin Rick Scherer - who designed and maintains a 25 host VMware ESX 3.5 farm with well over 300 Virtual Machines, plus writes a great blog (http://vmwaretips.com/wp/) - will be joining us to describe how his team uses SMVI. There will also be a panel of folks, including best practices authors and reference architects, to address questions submitted via chat.

Webcast: http://communications.netapp.com/p/Network_Appliance/20090219140000WL?REF_SOURCE=communities

Lisa Haut-Mikkelsen

SMVI Product Manager

[amiller_1]

Absolutely YES on the P.S. comment -- I'd love to see a central "SnapManager" console with agents that go into each application/application server. Especially with the new Bundles (which are great overall), it's more common now for people to want to implement SnapManager for Exchange + SQL + VMware + Sharepoint...the current installation process is frankly a bit tedious (not even going into the day to day maintenance/tracking).

Now....I realize this is the approach another storage vendor takes....but I still think there's some merit to the idea.

[amritad]

Hi

Recommended course of action for permanent resolution:

• In SMVI, delete all existing backups of the VM
  
• Go into the filer, and delete all snapshots of that volume which have SMVI in the name
  
• Go into SMVI, delete the existing backup job
  
• In SMVI, create a new backup job with a different name than the old one
  
• Re-run the new backup job to verify it is working
  

To test that this is in fact the issue and you are not encountering another problem, log into FilerView and delete the snapshot named _recent. Once you have tested the SMVI backup is functioning after deleting the existing _recent snapshot, please follow the steps above to properly and permanently resolve the issue and prevent further problems with SMVI backups.

IS this not working?

Regards

Amrita

[amritad]

Hi

Thanks for the great feedback. Please keep it coming. It will help us make our product better

REgards

Amrita

[selicsnetapp]

Hello,

I have SMVI installed on different servers. One running windows server 2003 r2 and windows server 2008 r2. My question is:

• I'm backing up SQL servers and one of them keeps timing out and the other one never has the problem. Do i need snapmanager for sql?
  
• when I get the logs back from my windows 2008 r2 server, I get them in this format https://[fe80:0:0:0:0:10 etc.. so I have to change the link to the server's ip to get the log. is this related to windows server 2008 r2?
  

p.s: in the future releases of this app, will it be possible to create profiles? At the moment, SMVI uses just one local profile.

Thank you in advance!

Regards.

[txskibum2000]

Im having an issue after a SMVI Restore on a particular volume, the Snapshots will show "0 KB" in FilerView and in System Manager afterwards.

[amritad]

So is the VM getting corrupted after restore? Do you see any errors in SMVI?

REgards

Amrita

[tirou]

Hi

I am trying to add the vFiler under SMVI and the vFiler is across the firewall. Using what protocol and port SMVI uses to communicate to the Filer?

What are the ports required to be opened at the firewall?

Thanks, Tirou

[amritad]

Hi

For SMVI, make sure the following ports are kept open:

8043: SMVI client uses this port to communicate with the SMVI server

By default, the SMVI server tries to communicate with the NetApp controller on port 443 using HTTPS. If HTTPS is not enabled, it will fall back to HTTP using port 80.

Regards

Amrita

[stuartwatkins]

Hi,

We are using SMVI v2 and one of the jobs is stuck in a running state and has been for a few days now. It is slowing filling up the free space on the VMFS LUN. When this fills up it will kill the VM which has the snap shots running (at the moment, the Job backs up all 4 VM's in the Datastore, and 3 of the VM's appear to have snapshorts created and so far not delete like in the normal SMVI process)

How do I stop the job (the GUI only gives me the option to suspend which doesn't help as that just stops the next job from running.

And what is the best way to "clean up" the left over Snapshots which are showing on the VM's in VIC.

If any more info is needed, just let me know.

Cheers in advance.

[collieritops]

Interesting -- I just found one of my SMVI jobs today stuck in a 'running' state. I restarted the SMVI service and it was still this way. suspend/resume was no help --  so I tried 'run job now' and got 2 alert emails..1 failure and 1 success, yet the job still shows running so I'm pretty confused.

Thoughts? Recommendations? How do you get the job out of the 'running' state? And please don't tell me that the answer is to delete the job and recreate it with a different name That was a poor solution for another SMVI issue we had.

[djaquays0]

I believe we had this happen in the past.  I believe we were instructed to stop the service, close the GUI and delete the contents of the \Program Files\NetApp\SMVI\server\crash folder.  I would probably wait for someone from NetApp to respond before doing so, as I'm not positive that is the correct folder.

[collieritops]

Looks like you're right... I just found a KB on it. Maybe it needs to be updated to apply to SMVI 2.x...it only mentions 1.0.

I was able to get around it by deleting and recreating the job with a new name (which changes our snap name, which also forces us to update our ndmp backup path in netbackup...ugh).

Solution ID: kb45786

Last updated: 26 MAY 2009

How to recover from a locked backup or restore job in SMVI

View Environment section

Symptoms

When using SnapManager for Virtual Infrastructure to backup or restore Virtual Machines, two operations may over lap that cause a lock to remain in place on a scheduled job. This job may remain in a 'running' state and any attempts to cancel the job will fail. Examination of the server logs show errors as the following: Attempting to re-run the job: ERROR - Backup <job name> is already in use. Please try this command again later. ERROR - FLOW-11019: Failure in AcquireLockWrapper: Backup <job name> is already in use. Please try this command again later. Canceling the job: ERROR - FLOW-11008: Operation failed: Backup <job name> is already in use. Please try this command again later. java.lang.RuntimeException: Backup <job name> is already in use. Please try this command again later.

Keywords : smvi, smvi locked, smvi back in use, FLOW-11019: Failure in AcquireLockWrapper,FLOW-11008: Operation failed

Cause of this problem

SnapManager for VI applies a lock on the backup set when performing operations such as a restore. It is possible that when performing a restore that a scheduled backup job will fire off. SnapManager for VI 1.0 should fail that job or at least delay it until after the restore completes. If for some reason there is a failure during the restore and that job never completes, it could lead to this condition.

Solution

SnapManager for VI utilizes an internal database to keep track of these locks and provides persistence across reboots. Simply rebooting the SnapManager for VI host will not clear these locks. If you want to remove all currently running tasks in SMVI, perform the following: Stop SnapManager for VI service. Remove the <SMVI dir>\server\crash directory. Start SnapManager for VI service. Performing these steps will not affect the scheduled jobs nor remove them from the interface. It will kill and remove any outstanding or in process tasks.

Environment

SnapManager 1.0 for Virtual Infrastructure

[collieritops]

We have a couple ESX datastores being backed up with SMVI and when they're backed up, the VMs in that datastore lose network connectivity briefly (a minute or less usually). This has been generating alerts from our monitoring software each night right when they backup. Has anyone else experienced this?

Also -- on one datastore that seems to do this the most consistently, one of the VMs it contains is the VirtualCenter server itself where SMVI runs. Is there a best practice recommendation on backing up the VirtualCenter/SMVI server using SMVI? I was thinking to get it out of the SMVI process and back it up via other means. I'm not sure if or how this would play into the network connection issues though.

[amritad]

Hi

Although SnapManager for Virtual Infrastructure is supported when running within a virtual machine, additional steps must be taken if SMVI will be “backing up itself” to avoid potential issues after a restore of the SMVI virtual machine. Note: SMVI “backing up itself” works for crash consistent snapshots. We recommend using the Windows native backup utility within the guest or some other backup application.

When the virtual machine running SnapManager for Virtual Infrastructure is backed up by SMVI, the current state of the SMVI backup workflow is captured within the Snapshot copies. Should the virtual machine be restored at a later date, SMVI within the virtual machine assumes that the host has failed in mid-workflow and attempts to resume backups at the point at which the Snapshot copy was taken. This can cause backups to run multiple times even if they were originally successful.

To prevent this issue, we recommend taking the following steps when SnapManager for Virtual Infrastructure is installed on a virtual machine:

1. Set the Startup Type of the SnapManager VI Windows Service to Manual. This will prevent SnapManager for Virtual Infrastructure from restarting automatically when the virtual machine it’s installed on is powered up after it has been restored.

2. After restoring the SMVI virtual machine, power the VM on and remove the contents of the %PROGRAMFILES%\NetApp\SMVI\server\crash directory. SMVI uses this directory after a crash to resume failed backups.

3. Start the SnapManager VI Windows Service. NetApp also recommends monitoring the backup of the virtual machine containing SnapManager for Virtual Infrastructure. Should timeouts occur during the VMware snapshot process, configure the backups to run

with the VMware snapshots disabled on the individual virtual machine.

Regards

Amrita

[collieritops]

I just read over at the Virtual Storage Guy blog that VSC 2.0 including the SMVI snap-in component has been announced and is scheduled for release & download on July 19th.

http://blogs.netapp.com/virtualstorageguy/2010/07/netapp-vsc-20-vcenter-plug-in-announced.html

Is this the new version of SMVI? What are the enhancements other than the obvious integration into a vCenter tab with VSC?

Are there licensing changes? In the comments I read "

Re: [NetApp - The Virtual Storage Guy] smccl submitted a comment to NetApp VSC 2.0 vCenter Plug-in Announced

@SMCCL Great question. The NetApp VSC, including SMVI the RCU is a free download to all customers. There are some capability restrictions if one does not have SnapRestore (with SMVI) and FlexClone (with RCU) licensed on the array, but both of these licenses are pretty popular and rather inexpensive."

Am I reading this correct that SMVI is a free download now?!

Please update us.

Thx

[amritad]

Hi

VSC 2.0 is our integrated plug-in consisting of VSC, Provisioning and cloning (RCU) and backup an recovery(SMVI). VSC and RCU are free. For RCU you would still need the FlexClone license for cloning. For SMVI you would still need the license ,it is not free.

Regards

Amrita

[collieritops]

Can you please provide more information?

What's new in this version besides the VSC/vCenter-tab integration?

[cnshelpdesk]

We like running NFS with 2TB volumes for dedupe/manageability/etc.  We can have anywhere from 10-50 powered on VMs per datastore

I know it would depend on how much I/O/load on ESX/etc is going on during backups but with SMVI 2.0 and grouping how many VMs in a datastore are people running?

I've heard people taking groups of 30... but then you're taking 3 snapshots a day out of your 255... (boy makes me also wonder.. when are they getting rid of the 255 snapshot limit!)

S

[imarcroft]

Hi All

wondered if anyone else has seen this problem

all SMVI notifications have links in, which they reference what looks like an IPv6 address in which doesn't work as we are using IP v4

it has the following format

htp://[fe80:0:0:0:84d8:47ef:8b30:f087%11]:8043/smvi/logViewer?id=backup_Daily 8.15 RDM's attached_20100706201500

the fe80:0:..... is the part I refer to, if I replace this with the netbios name the links work

this is the same for the .sfr file for single file recovery, if you edit the .sfr file it too references the fe80:00 number

when I look at the IP stack for the card of the VC server IPv6 was selected, I have deselected IPv6 and rebooted the VC server. but makes no difference.

any help much appreciated

thanks

[nlfmmeijburgienasb]

Hello,

I have the same problem with ipv6 address in the link to the logfile.

IPv6 allready removed from server, server rebooted.

Anyone a clue?

Arnold

[imarcroft]

Hi The only way we could fix this was by removing SMVI completely, make sure IPv6 isn't bound to the network cards.

We reinstalled with the version2.0 of virtual storage console.

good luck