SR_BACKEND_FAILURE_46 The VDI is not available

[Daniele Renda1709154949]

Hi guys,

on my Xen server 6.1 I'm trying to remove an old vdi that I'm not using in none VM.

 

So I tried first to remove the VDI of 300Gb:

 

xe vdi-destroy uuid=80b8d9af-705c-431e-b1cb-66a835de7479
Error code: SR_BACKEND_FAILURE_46
Error parameters: , The VDI is not available [opterr=Command ['/usr/bin/vhd-util', 'query', '--debug', '-vsf', '-n', '/dev/VG_XenStorage-58d108cc-e9d9-fa6d-e55a-6f53fdd0b786/VHD-80b8d9af-705c-431e-b1cb-66a835de7479'] failed (22): error opening /dev/VG_XenStorage-58d108cc-e9d9-fa6d-e55a-6f53fdd0b786/VHD-80b8d9af-705c-431e-b1cb-66a835de7479: -22],
 

 

Then I tried to forget the vdi:

xe vdi-forget uuid=80b8d9af-705c-431e-b1cb-66a835de7479

 

and then rescan the storage:

xe sr-scan uuid=58d108cc-e9d9-fa6d-e55a-6f53fdd0b786
Error code: SR_BACKEND_FAILURE_46
Error parameters: , The VDI is not available [opterr=Error scanning VDI 80b8d9af-705c-431e-b1cb-66a835de7479],
 

So now I have these 300Gb busy also if the vdi is not more present. How I can to definitly remove these disk?

 

Thanks so much!

 

 

[James Cannon]

The VDI forget only purges entry from XenServer database. So, the VDI-destroy is failing for same reason as SR scan. Can you change pool master? Maybe reboot master if you can migrate VMs off to other host.

[Daniele Renda1709154949]

I can reboot the physical machine during night if this is needed. This should solve the problem?

[Daniele Renda1709154949]

I just made a test stopping all Vm and restard the physical machine but the problem persist.

 

To tell the thruth the vdi that now is not more listed (80b8d9af-705c-431e-b1cb-66a835de7479) is a disk that I want to delete. I'd like to remove all space allocation for this disk that now I don't see anymore in the XenCenter.

 

How I can solve this problem?

 

Thanks!

[James Cannon]

IF VDI is not linked to other VDI, you can try to use lvrmove (for LVM-based VDI) or rm (for .vhd on file system). Once done, scan SR.

[Daniele Renda1709154949]

Thanks for the suggestion, I done

 

lvdisplay -v
    Finding all logical volumes
 

    ...

    ...

      --- Logical volume ---
  LV Name                /dev/VG_XenStorage-58d108cc-e9d9-fa6d-e55a-6f53fdd0b786/VHD-80b8d9af-705c-431e-b1cb-66a835de7479
  VG Name                VG_XenStorage-58d108cc-e9d9-fa6d-e55a-6f53fdd0b786
  LV UUID                WjZHQy-OrU3-1cUN-OSW3-LPs0-tE36-s7WWDt
  LV Write Access        read/write
  LV Status              NOT available
  LV Size                300.59 GB
  Current LE             76952
  Segments               1
  Allocation             inherit
  Read ahead sectors     auto
 

then:

 

 lvremove /dev/VG_XenStorage-58d108cc-e9d9-fa6d-e55a-6f53fdd0b786/VHD-80b8d9af-705c-431e-b1cb-66a835de7479
  Logical volume "VHD-80b8d9af-705c-431e-b1cb-66a835de7479" successfully removed
 

then I retried to rescan the SR:

 

xe sr-scan uuid=58d108cc-e9d9-fa6d-e55a-6f53fdd0b786
Error code: SR_BACKEND_FAILURE_40
Error parameters: , The SR scan failed  [opterr=Command ['/usr/sbin/lvchange', '-ay', '/dev/VG_XenStorage-58d108cc-e9d9-fa6d-e55a-6f53fdd0b786/VHD-80b8d9af-705c-431e-b1cb-66a835de7479'] failed (5): One or more specified logical volume(s) not found.],
 

 

What is the problem now? Seems that Xen has the reference to VHD-80b8d9af-705c-431e-b1cb-66a835de7479 somehwere but I don't understand where!

 

Thanks!

[Daniele Renda1709154949]

Someone can help me please?

 

Thanks

[James Cannon]

You can try to remove the MGT volume ...

 

lvdisplay | grep MGT

 

Once done, find the one that matches the UUID of your SR and then use lvremove for the MGT logical volume, rescan the SR and you should be fine. Issue now is reference to logical volume in MGT header. 

[Daniele Renda1709154949]

Hi, I can remove the MGT also when all vm are active?

 

Thanks

[James Cannon]

Yes. 

[Daniele Renda1709154949]

It worked perfectly!! Thanks for your hints!

[James Cannon]

Glad to hear all is well.

[Aaron Murray1709160777]

Hope I can bring this one back to life.

I've been having a somewhat similar issue to the OP.

 

Server did a hard shutdown.  On reboot, one of the VMs wasn't running.  I detached two recently added disks, and the VM ran ok.

Looking at the SR, if I re-scanned, I got an error VDI not available.

 

I tried forgetting the VDIs etc and then re-scanning.  on the CLI, this gave SR_BACKEND_FAILURE_46

 

I followed the advice above to use lvdisplay , then lvremove.

No errors showing now, but if I do a rescan, I get nothing.  Are they gone?

If I start over, is there some way to stop it happening again?

 

Cheers