VMware Cloud Community
StephaneZ
Contributor
Contributor

ESXi Mounting another drive VMFS partition ?

I have a VMFS partition that I need to mount in my ESXi server from another system that just died.

I keep reading that ESXi should try and mount any VMFS partition detected. But it is not mounting this one.

The SATA drive is being detected in the "Storage Adapters", but when I try to add "Storage", every time I try and add it, it warns me that it will re-format all data on the partition. I just want it to mount the existing partition that shows up as a VMFS partition type.

The drive appears fine the esxcfg-mpath and shows as I would expect, but the partition is not being mounted?

Is there a way to force the mount so that I can recover the data from this partition?

Stephane

Reply
0 Kudos
13 Replies
jbruelasdgo
Virtuoso
Virtuoso

in the configuration tab, go to Storage Adapters

and click in the Rescan... link, select Scan for new VMFS Volumes

do not add it as you are trying, because if you do so, it will format all the data

regards

Jose Ruelas

Jose B Ruelas http://aservir.wordpress.com
Reply
0 Kudos
depping
Leadership
Leadership

Check the log files, you might need to resignature the disk before you can actually mount it!

Duncan

VMware Communities User Moderator

-


Blogging:

Twitter:

If you find this information useful, please award points for "correct" or "helpful".

Reply
0 Kudos
StephaneZ
Contributor
Contributor

I have done this multiple times, nothing get's added. The VMFS partition I am trying to mount was part of an extend. I'm not sure if that might have to do with it.

Reply
0 Kudos
StephaneZ
Contributor
Contributor

I have enabled the LVM re-signature flag in the Adv. Config, nothing jumps out as wrong.

Do I need to do something specific to have it re-sign?

Reply
0 Kudos
depping
Leadership
Leadership

What do you mean by "part of an extend"? You normally can't break an extend and use it again....

If you enabled the resignaturing, rescan for new luns again and disable it and check if it shows up.

Duncan

VMware Communities User Moderator

-


Blogging:

Twitter:

If you find this information useful, please award points for "correct" or "helpful".

StephaneZ
Contributor
Contributor

I enabled resignature, did a rescan for new luns, as suggested and it does not show up.

What logs should I be looking at?

Reply
0 Kudos
Dave_Mishchenko
Immortal
Immortal

/var/log/messages

You mention extents above, was this drive added as an extent to another datastore?

StephaneZ
Contributor
Contributor

Hi, Dave.

We had an ESXi server fail last weekend, the partition that I am trying to move to this ESXi server was the first partition of an extent with two partitions.

What seems to be the case is that since the two partitions in the extent were on a different RAID adapter, the vmhdX:X:X don't match and it can't seem to detect it.

Any suggests would be reallly helpful.

Stéphane

Reply
0 Kudos
StephaneZ
Contributor
Contributor

Here is an entry from the /var/log/messages showing it detected the drive:

May 4 18:34:26 vmkernel: 0:00:00:12.615 cpu0:1244)SCSI: 861: GetInfo for adapter vmhba1, , max_vports=0, vports_inuse=0, linktype=0, state=0, failreason=0, rv=-25, sts=bad0001

May 4 18:34:26 vmkernel: 0:00:00:12.615 cpu0:1244)ScsiScan: 395: Path 'vmhba1:C0:T0:L0': Vendor: 'ATA ' Model: 'WDC WD6400AAKS-7' Rev: '01.0'

May 4 18:34:26 vmkernel: 0:00:00:12.615 cpu0:1244)ScsiScan: 396: Type: 0x0, ANSI rev: 5

May 4 18:34:26 vmkernel: VMWARE SCSI Id: Supported VPD pages for vmhba1:C0:T0:L0 : 0x0 0x80 0x83

May 4 18:34:26 vmkernel: VMWARE SCSI Id: Device id info for vmhba1:C0:T0:L0: 0x2 0x0 0x0 0x18 0x4c 0x69 0x6e 0x75 0x78 0x20 0x41 0x54 0x41 0x2d 0x53 0x43 0x53 0x49 0x20 0x73 0x69 0x6d 0x75 0x6c 0x61 0x74 0x6f 0x72

May 4 18:34:26 vmkernel: VMWARE SCSI Id: Id for vmhba1:C0:T0:L0 0x20 0x20 0x20 0x20 0x20 0x57 0x44 0x2d 0x57 0x43 0x41 0x53 0x59 0x31 0x31 0x39 0x38 0x30 0x33 0x31 0x57 0x44 0x43 0x20 0x57 0x44

May 4 18:34:26 vmkernel: 0:00:00:12.616 cpu2:1244)ScsiUid: 653: No supported identifier types found in VPD device id page.

May 4 18:34:26 vmkernel: 0:00:00:12.616 cpu2:1244)ScsiScan: 559: Path 'vmhba1:C0:T0:L0': No standard UID: Failure

May 4 18:34:26 vmkernel: 0:00:00:12.616 cpu2:1244)ScsiScan: 641: Discovered path vmhba1:C0:T0:L0

May 4 18:34:26 vmkernel: 0:00:00:12.616 cpu0:1244)SCSI: 1953: Device vmhba1:0:0 has not been identified as being attached to an active/passive SAN. It is either attached to an active/active SAN or is a local device.

May 4 18:34:26 vmkernel: 0:00:00:12.616 cpu0:1244)ScsiPath: 3178: Plugin 'legacyMP' claimed path 'vmhba1:C0:T0:L0'

May 4 18:34:26 vmkernel: 0:00:00:12.633 cpu0:1244)ScsiDevice: 3643: Adding Device "vml.0100000000202020202057442d574341535931313938303331574443205744" from Plugin "legacyMP", Device Type 0

May 4 18:34:26 vmkernel: 0:00:00:12.633 cpu0:1244)SCSI: 5498: Logical device vml.0100000000202020202057442d574341535931313938303331574443205744 for target vmhba1:0:0 was registered successfully.

May 4 18:34:26 vmkernel: 0:00:00:12.633 cpu0:1244)SCSI: 861: GetInfo for adapter vmhba1, , max_vports=0, vports_inuse=0, linktype=0, state=0, failreason=0, rv=-25, sts=bad0001

I don't see anything specific that shows it is having problems mounting the partition...

Reply
0 Kudos
StephaneZ
Contributor
Contributor

Also looks like fdisk sees the drive and partition too:

Disk /dev/disks/vmhba1:0:0:0: 640.1 GB, 640133946880 bytes

255 heads, 63 sectors/track, 77825 cylinders

Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System

/dev/disks/vmhba1:0:0:1 1 18114 145500641 fb VMFS

Reply
0 Kudos
StephaneZ
Contributor
Contributor

Well... I opened a support request with VMware to resolve the problem.

I must say, they are pretty good at what they do! I'm back up and running. Thanks for all the help.

Reply
0 Kudos
Dave_Mishchenko
Immortal
Immortal

So what did you end up having to do and did you mount both the original drive for the datastore and all the extents you added?

Reply
0 Kudos
StephaneZ
Contributor
Contributor

The way VMware handles the extents is by using LVM. When you move an extent between systems or RAID adapters, the pointers between the extents is miss-represented.

So when you do a exscfg-mpath it returns that one of the extents are missing, in my case nothing to re-scan will ever find them because the extent is looking for a disk on vmhba2:0:0:2, and the drive/partition was on vmhba1:0:0:1.

Using dd, the LVM pointer to the new location was set. Not something for the faint of heart Smiley Happy But the VM engineer's that did the work sure seemed to know what was going on Smiley Happy

I'm just grateful for a company who actually has the engineering and support capability to actually get my storage issues resolved. Now if only I could figure out why my e1000 doesn't transfer at anything more then 80Mbps.

Note to self: NEVER USE EXTENTS!!!!

Stéphane

Reply
0 Kudos