sammyxes
Contributor
Contributor

Please help: extent missing from datastore

Jump to solution

Hi,

I have a datastore made of a 20TB disk and an extent of about 40TB (local disks, areca raid controller).

Long story short, I had to delete and re-create the 40TB volume on the RAID controller (without initalization, of course).

Now all the data is there but eui changed for the 40TB volume and vmware is unable to mount it.

log/hostd.log:2016-05-10T07:17:16.107Z info hostd[3D380B70] [Originator@6876 sub=Vimsvc.ha-eventmgr] Event 118 : An attached device eui.001b4d2051676528:1 may be offline. The file system [backvol1, 55e980d4-386dfa7f-7cb2-0cc47a09ba36] is now in a degraded state. While the datastore is still available, parts of data that reside on the extent that went offline might be inaccessible.

[root@esxi2:/vmfs/volumes/55e980d4-386dfa7f-7cb2-0cc47a09ba36] esxcli storage vmfs snapshot list

   Volume Name:

   VMFS UUID:

   Can mount: false

   Reason for un-mountability: some extents missing

   Can resignature: false

   Reason for non-resignaturability: some extents missing

   Unresolved Extent Count: 1

Can anyone help me please?

Thank you!

0 Kudos
1 Solution

Accepted Solutions
continuum
Immortal
Immortal

Hi
you asked for instructions for the procedure to readd a missing extend.
Basically thats quite easy:
in the VMFS header section of the base VMFS-volume you can assign the extends.
Lets say the base-volume is mpx.vmhba1:C0:T3:L0:1
first extend is mpx.vmhba1:C0:T4:L0:1
second extend is mpx.vmhba1:C0:T5:L0:1
Then you will see this strings in the VMFS-header:
vmhba1:3:0
and a bit later

mpx.vmhba1:C0:T3:L0:1

mpx.vmhba1:C0:T4:L0:1

mpx.vmhba1:C0:T5:L0:1
You just have to edit the list of extends.
But: normally you can not edit this section at all and to add to the confusion this values are buffered in RAM so it is not trivial to change this values correctly.
I highly recommend that instead of trying this yourself and repeating all the mistakes I made while learning this - just call me.
I rather assist you personally than give dangerous advice that will likely make matters worse.


________________________________________________
Do you need support with a VMFS recovery problem ? - send a message via skype "sanbarrow"
I do not support Workstation 16 at this time ...

View solution in original post

0 Kudos
11 Replies
Marmotte94
Enthusiast
Enthusiast

hi,

Can you make "Rescan all" VMHBA

> Add lun

and choose "Keep Existing Signature or Assign a New Signature" When you see the prompt message.

Thank you

Regards,

Please, visit my blog http://www.purplescreen.eu/
0 Kudos
sammyxes
Contributor
Contributor

I can't do that, I don't have the option to keep the existing signature (it's an extent, not a volume)

vm.png

0 Kudos
Marmotte94
Enthusiast
Enthusiast

Hi,

Do you try mount VMFS Volume with this command ?

#esxcli storage vmfs snapshot mount -u uuid

Thank you,

Regards,

Please, visit my blog http://www.purplescreen.eu/
0 Kudos
sammyxes
Contributor
Contributor

Nothing...

I've deleted the volume, extents are bad Smiley Sad

0 Kudos
continuum
Immortal
Immortal

Looks like I am coming too late ?
Extends that have changed their UUIDs can be reattached by hexediting the parent LUN. I am not sure if this is documented anywhere but it is possible.
Ulli


________________________________________________
Do you need support with a VMFS recovery problem ? - send a message via skype "sanbarrow"
I do not support Workstation 16 at this time ...

0 Kudos
sammyxes
Contributor
Contributor

Too late, but thanks anyway.

Can you provide the details for hex-editing, in case anyone else gets here?

Thank you!

0 Kudos
continuum
Immortal
Immortal

Hi
you asked for instructions for the procedure to readd a missing extend.
Basically thats quite easy:
in the VMFS header section of the base VMFS-volume you can assign the extends.
Lets say the base-volume is mpx.vmhba1:C0:T3:L0:1
first extend is mpx.vmhba1:C0:T4:L0:1
second extend is mpx.vmhba1:C0:T5:L0:1
Then you will see this strings in the VMFS-header:
vmhba1:3:0
and a bit later

mpx.vmhba1:C0:T3:L0:1

mpx.vmhba1:C0:T4:L0:1

mpx.vmhba1:C0:T5:L0:1
You just have to edit the list of extends.
But: normally you can not edit this section at all and to add to the confusion this values are buffered in RAM so it is not trivial to change this values correctly.
I highly recommend that instead of trying this yourself and repeating all the mistakes I made while learning this - just call me.
I rather assist you personally than give dangerous advice that will likely make matters worse.


________________________________________________
Do you need support with a VMFS recovery problem ? - send a message via skype "sanbarrow"
I do not support Workstation 16 at this time ...

0 Kudos
sammyxes
Contributor
Contributor

Well, as I said before, I already deleted the datastore (I've made a clone of the VM on another volume before fiddling with this one and now I'm moving the VM back).

The thing is that the path to the LUN was the same (vmhba3:C0:T0:L2), only the device EUI was different, so I don't think that this was the correct answer in my case.

0 Kudos
continuum
Immortal
Immortal

I only do this type of edits in a recovery context.
After my edits I extract all I need and then I wipe the volume and create a new VMFS from scratch.


________________________________________________
Do you need support with a VMFS recovery problem ? - send a message via skype "sanbarrow"
I do not support Workstation 16 at this time ...

0 Kudos
ziffusion
Contributor
Contributor

continuum

I found your answer here from way back. Would you mind taking a look at a problem that I am having at this time? I would really appreciate it.

Datastore disappeared after disk replacement in a RAID 1 array (ESXi 6.7)

0 Kudos
SpitzerG
Contributor
Contributor

Hi, I have exact the same Problem, my extend has a new id now, how can i change the vmfs header correctly, download and edit in hex editor, what else I have to take into consideration? thx

0 Kudos