VMware Cloud Community
ajnrock
Enthusiast
Enthusiast

Can't Power Up a VM - A File Was Not Found

I have a VM that has been working without issue now for about six months. Last night I go to migrate to another host cold, shut it down, do a run of the mill migrate, and then power it back on. When I power on it initializes the power on, and then gives the error "A File Was Not Found" I can still successfully see all of the Guests Virtual Disks in the Datastore Browser, however the one thing that is a bit odd is that the datastore browser is very slow only when I open this specific VM in this Datastore. All other VMs on same datastore continue to function per normal.

I have also tried migrating back to original host as well as powering up through the host instead of vcenter as a last resort. All with the same result.

We are running ESX 3.5 & VCenter 2.5, configured for HA.

Thanks for any thoughts on what else I could try without kicking the hosts.

Adam

Reply
0 Kudos
12 Replies
oreeh
Immortal
Immortal

Post a listing of the files in the VM folder, the VMX file and the latest logfile.

Reply
0 Kudos
VMKR9
Expert
Expert

edit the virtual machine properties and confirm that the path to the vmdk files are correct, you could try removing the disks and adding them back in

Reply
0 Kudos
ajnrock
Enthusiast
Enthusiast

Here they are as attachments to spare everyone a ten mile long log file in the post.

Reply
0 Kudos
demz
Expert
Expert

Unregister your VM from the VC, check the disk path in the .vmx of your VM via the COS, if there is a mistake correct it and then re-register your VM in order to power it on.

If you don't find any issue in the vmx, please post it right here.

Hope this helps.

Reply
0 Kudos
oreeh
Immortal
Immortal

Modify the VMX and correct the incorrect path "/vmfs/volumes/468847f8-08fa9ca3-1f00-0018fe72bf90/CL-SQLSP-N01/CL-SQLSP-N01_3.vmdk"

Reply
0 Kudos
demz
Expert
Expert

Oreeh, the king of the vmware.log Smiley Wink

Reply
0 Kudos
ajnrock
Enthusiast
Enthusiast

Well, I think I got all of the cobwebs out of my head by now, but I am not seeing anything wrong with the path, and the file is there. I have looked at this several times. Am I completely missing the boat?

Reply
0 Kudos
oreeh
Immortal
Immortal

According to the logfile

May 23 06:13:37.951: vmx| DISKLIB-FLAT_VMFS  :Setting multi-writer flagDISKLIB-VMFS : "/vmfs/volumes/468847f8-08fa9ca3-1f00-0018fe72bf90/CL-SQLSP-N01/CL-SQLSP-N01_3-rdmp.vmdk" : failed to open (25): AIOMgr_Open failed. Type 10
May 23 06:13:37.982: vmx| DISKLIB-DSCPTR: Failed to open extents for descriptor file in normal mode
May 23 06:13:37.983: vmx| DISKLIB-LINK  : "/vmfs/volumes/468847f8-08fa9ca3-1f00-0018fe72bf90/CL-SQLSP-N01/CL-SQLSP-N01_3.vmdk" : failed to open (The system cannot find the file specified).  
May 23 06:13:37.983: vmx| DISKLIB-CHAIN : "/vmfs/volumes/468847f8-08fa9ca3-1f00-0018fe72bf90/CL-SQLSP-N01/CL-SQLSP-N01_3.vmdk" : failed to open (The system cannot find the file specified).
May 23 06:13:37.983: vmx| DISKLIB-LIB   : Failed to open '/vmfs/volumes/468847f8-08fa9ca3-1f00-0018fe72bf90/CL-SQLSP-N01/CL-SQLSP-N01_3.vmdk' with flags 0x4a (The system cannot find the file specified).
May 23 06:13:37.983: vmx| Msg_Post: Error
May 23 06:13:37.983: vmx| [http://msg.disk.fileNotFound|http://msg.disk.fileNotFound] VMware ESX Server cannot find the virtual disk "/vmfs/volumes/468847f8-08fa9ca3-1f00-0018fe72bf90/CL-SQLSP-N01/CL-SQLSP-N01_3.vmdk". Please verify the path is valid and try again.
May 23 06:13:37.983: vmx| [http://msg.disk.noBackEnd|http://msg.disk.noBackEnd] Cannot open the disk '/vmfs/volumes/468847f8-08fa9ca3-1f00-0018fe72bf90/CL-SQLSP-N01/CL-SQLSP-N01_3.vmdk' or one of the snapshot disks it depends on.
May 23 06:13:37.983: vmx| [http://msg.disk.configureDiskError|http://msg.disk.configureDiskError] Reason: The system cannot find the file specified.----------------------------------------

ESX is not able to access /vmfs/volumes/468847f8-08fa9ca3-1f00-0018fe72bf90/CL-SQLSP-N01/CL-SQLSP-N01_3.vmdk.

And since the first line mentions CL-SQLSP-N01_3-rdmp.vmdk I assume this is a RDM and the storage in question isn't presented to ESX or the ESX doesn't have access to it (for whatever reason).

ajnrock
Enthusiast
Enthusiast

That is in fact a RDM, but since no changes have been made to the guest config other than the host migration, I am going to focus on the SAN paths for a bit. I have been staring looking for a syntax error, but if as you say the storage is just not being presented, that might make sense. These RDMs are to shared LUNs for a MSCS Cluster, but this guy has been up and running for months without issue, so I am really scratching my head. I might do a storage rescan in a bit and see if that changes anything and have my storage guy give the LUN a once over.

I will report back. thanks for the effort!

Reply
0 Kudos
ajnrock
Enthusiast
Enthusiast

Just to close the loop on this thread - It would seem that I can not have 2 VMs on the same host pointed at the same RDM in Physical Compatability Mode. I spent quite a bit of time playing with this and the only thing I could do to recover is run the guests on seperate physical hosts in the VMWare HA Cluster.

Thanks anyhow for the efforts.

Adam

Reply
0 Kudos
cgdii
Contributor
Contributor

I am having the same issue, but I only have one ESX Host I can run these VMs on. I am attempting to set up a Veritas Cluster using RDMs for training purposes and if I power them on while they are on the same ESX Host I get this error in the vmware.log:

Jul 23 21:07:46.459: vmx| DISKLIB-LIB : Opened "/vmfs/volumes/48121a7e-db8f5588-02c6-001ec9353f72/PC2TRAIN02/PC2TRAIN02.vmdk" (flags 0xa). 0x8787c6c

Jul 23 21:07:46.478: vmx| DISK: OPEN '/vmfs/volumes/48121a7e-db8f5588-02c6-001ec9353f72/PC2TRAIN02/PC2TRAIN02.vmdk' Geo (4699/255/63) BIOS Geo (0/0/0) freeSpace=142019Mb, lsilogic

Jul 23 21:07:46.480: vmx| DISK: OPEN scsi1:0 '/vmfs/volumes/48121a7e-db8f5588-02c6-001ec9353f72/PC2TRAIN02/PC2TRAIN02_1.vmdk' persistent R[(null)]

Jul 23 21:07:46.484: vmx| DISKLIB-VMFS : "/vmfs/volumes/48121a7e-db8f5588-02c6-001ec9353f72/PC2TRAIN02/PC2TRAIN02_1-rdm.vmdk" : failed to open (25): AIOMgr_Open failed. Type 10

Jul 23 21:07:46.484: vmx| DISKLIB-DSCPTR: Failed to open extents for descriptor file in normal mode

Jul 23 21:07:46.484: vmx| DISKLIB-LINK : "/vmfs/volumes/48121a7e-db8f5588-02c6-001ec9353f72/PC2TRAIN02/PC2TRAIN02_1.vmdk" : failed to open (The system cannot find the file specified).

Jul 23 21:07:46.484: vmx| DISKLIB-CHAIN : "/vmfs/volumes/48121a7e-db8f5588-02c6-001ec9353f72/PC2TRAIN02/PC2TRAIN02_1.vmdk" : failed to open (The system cannot find the file specified).

Jul 23 21:07:46.484: vmx| DISKLIB-LIB : Failed to open '/vmfs/volumes/48121a7e-db8f5588-02c6-001ec9353f72/PC2TRAIN02/PC2TRAIN02_1.vmdk' with flags 0xa (The system cannot find the file specified).

Jul 23 21:07:46.484: vmx| Msg_Post: Error

Jul 23 21:07:46.484: vmx| http://msg.disk.fileNotFound VMware ESX Server cannot find the virtual disk "/vmfs/volumes/48121a7e-db8f5588-02c6-001ec9353f72/PC2TRAIN02/PC2TRAIN02_1.vmdk". Please verify the path is valid and try again.

Jul 23 21:07:46.484: vmx| http://msg.disk.noBackEnd Cannot open the disk '/vmfs/volumes/48121a7e-db8f5588-02c6-001ec9353f72/PC2TRAIN02/PC2TRAIN02_1.vmdk' or one of the snapshot disks it depends on.

If I migrate one of the cluster nodes to another ESX Host, I can power it on and have no issues. Any thoughts?

Reply
0 Kudos
ajnrock
Enthusiast
Enthusiast

I did not pursue any further, generally we wanted on separate hosts under normal circumstances, and I have since added a third host to my VMWare Cluster so I really no longer have a need.

After all of my research, my take is that this is not possible currently in Physical Comparability Mode, at least over the same HBAs (WWNs). I was not able to find a definitive answer on this.

It may likely work in virtual comparability mode, I have not tried that.

Reply
0 Kudos