Direct SAN access failure troubleshooting

[ashrid]

Having installed VCB and run vcbsandbg my issue would appear to be this:

http://sysadmkb.blogspot.se/2011/10/vmw ... iscsi.html

I intend to discuss this with VMWare support before proceeding and appreciate my next question is off-topic but has anyone got any experience with the fix proposed in that linked page?

[habibalby]

Hi,
Do you face the same symptoms in the sphere client, that you lose the VMFS signature? I haven't notice this before, but it worth checking that out.

I moved from direct SAN to HotAdd, when I as on direct San, most of the time it failed, or it goes via network, it hardly ever works. With HotAdd, processing rate performance far than when I was on direct San. Average rate for full backup 60-80mb/s, incremental backup it gose from 300 - 1Gb and one job always incremental reaches 2Gb.

Try HotAdd you might see it works better than direct San access.

[Gostev]

By the way, please note that VCB does not support or "understand" VMFS-5, so vcbsandbg is no longer suitable as a troubleshooting tool these days.

[ashrid]

Hi,

Yes the sphere client shows no signature exactly as described in the linked article, all of my volumes as VMFS 3

[mbutch]

This thread was VERY helpful. I was getting SAN messages, so I checked the backup proxy. For some reason it was set on manual, when for our environment automatic should work fine. And it had old LUNs there anyway! So I switched it and we've been working fine.

[spgsit5upport]

Code: Select all

10/05/2013 20:41:13 :: Unable to establish direct connection to the shared storage (SAN).
Please ensure that:
- HBA is properly installed in the Veeam Backup server computer, or software iSCSI initiator is configured correctly.
- SAN volume can be seen by operating system in the Windows Disk Management snap-in on the Veeam Backup server.
- Read access is allowed for the Veeam Backup server computer on the corresponding LUN (refer to your SAN documentation).

10/05/2013 20:41:18 :: Direct SAN connection is not available, failing over to network mode...

I made absolutely SURE that Veeam server has access the LUNs
I can see each & every LUN in Disk Management on this server

The volumes it complains about are all virtual compatibility RDM on Dell MD3600i

Anybody has an idea how to tackle it?

Seb

[chrisdearden]

Have you done a storage rescan of the proxy server ? From the backup infrastructure tab right click the server under windows servers and select refresh.

[spgsit5upport]

Sure I did...

Sadly it is not possible to see in Veeam GUI what Veems can "see" itself

[veremin]

First of all, can you confirm that you've successfully followed steps required for proper configuring Direct SAN access that are described here.

If so, kindly see the existing discussion which is primarily concerned with Direct SAN troubleshooting stuff.

Hope this helps.
Thanks.

[spgsit5upport]

Ofcourse I have, as the normal vmdk DO get backup by by SAN

I have a server with 9 disks attached (3 disks are plain vmdk on VMFS volume which DO get backup by Direct SAN access, the remaining vRDM are done by network, even ALL are accessible to the Veeam server)

[Gostev]

If you have exhausted all troubleshooting tips listed in this topic, please open a support case and let our support engineers take a look at the debug logs. Please include support case ID and/or resolution for the future reference. Thanks!

[nmace]

I've got a file server VM with it's data stored in a VMDK that is on a dedicated SAN volume that is presented to the VM as a local drive via the Hypervisor (vSphere 4.1). I have several volumes setup and working successfully with the SAN Direct method. The data volume is setup the same way as the others (access is granted to the LUN via Veeam's iSCSI address, the volume is mounted under the iSCSI initiator, and is marked as "offline" in the Veeam server's Disk Management snap-in). However when Veeam gets to that section of the VM it fails over to the network method. This is a somewhat large VMDK (700+ GB). I'd very much like to back it up over my 10 gigabit links that Veeam has access to. Any ideas? I've setup several volumes like this, they all worked except this one. But I can't see where the problem is.

Veeam is installed on a physical server running Server 2012. Veeam is version 6.5.0.109.

[foggy]

Please review recommendations given in the topic you've been merged to. If nothing described above helps, feel free to ask for further assistance.

[nmace]

I've got a file server VM with it's data stored in a VMDK that is on a dedicated SAN volume that is presented to the VM as a local drive via the Hypervisor (vSphere 4.1). I have several volumes setup and working successfully with the SAN Direct method. The data volume is setup the same way as the others (access is granted to the LUN via Veeam's iSCSI address, the volume is mounted under the iSCSI initiator, and is marked as "offline" in the Veeam server's Disk Management snap-in). However when Veeam gets to that section of the VM it fails over to the network method. This is a somewhat large VMDK (700+ GB). I'd very much like to back it up over my 10 gigabit links that Veeam has access to. Any ideas? I've setup several volumes like this, they all worked except this one. But I can't see where the problem is.

Veeam is installed on a physical server running Server 2012. Veeam is version 6.5.0.109.

Actually I get this corrected on my own. I had two volumes the same size on the SAN and had mistakenly setup Veeam with access to Volume A and was trying to backup Volume B. I corrected the issue and it worked like a charm. Thanks anyway though.

[foggy]

Glad you've nailed it, thanks for getting back.