Quantum Forum for DXi V5000
VM was not found in /export
I have 5 Linux VMs which are backed up daily via vmpro to a dxi. Several days ago one of our IT guys did something dumb and none of the VMs could be backed up anymore. That has since been resolved, but 2 of the VMs still can't be backed up.
When I run a backup on the folder I get the following message:
"VM was not found in /export, check logs for more details."
Looking at the logs I found the following entry in vm_proxy_fs:
2013-12-03 10:04:09.455192: WARNING [thread:140590117287680, where:vmxtobasevmdk.c:224, error:5] [FataDataStore] rails/rails.vmx: 500:Internal Server Error: Input/output error: Input/output error
vm_proxy_fs[3040]: WARNING [thread:140590117287680, where:vmxtobasevmdk.c:224, error:5] [EVA3000%2520Datastore] quoteshare/quoteshare.vmx: 500:Internal Server Error: Input/output error: Input/output error
I looked on the /export file system and I found that the VMs are in fact, there. However the .vmx files are not present for these 2 VMs in their respective directories.
Just for grins I took one of the VMs and put it into a new folder I created in vmpro. I then created a policy for it and attempted to run a backup. Same results.
I find myself going in circles now so I'm throwing this out there to see if anyone has any suggestions.
thanks
Replies to This Discussion
-
Permalink Reply by Jerry Schmidt on
-
Hello Robert,
For the two VMs still giving you issues, check the dates of the vmdk files that are listed in /export, are they current? If they are not, select the VMs from the the vmPro GUI, and deselect export and CBT. Power cycle vmPro, select the VMs again, enable export and cbt, then check /export to see if the vmdk and vmx files pull through correctly.
Even if the vmdk files are current you can try the above steps to see if this allows the vmx files to appear in /export.
-
Permalink Reply by Robert Smith on
-
Jerry, thanks for the suggestions. The dates on the files were current, but I un-exported them, rebooted vmpro, and re-exported just to see if it would help. I'm still getting the same problem. I checked the files this morning and they are all dated 12/5/2013, which would be correct and it matches the VMs that are working correctly. Still no .vmx files in the 2 affected VMs though.
-
-
If you check the VMs via vmWare, do the vmx and vmdk files reside on the same datastore?
Try deselecting export on the VMs again, reboot and then re-export the VMs. Then via the vmPro GUI go to help--> vmPro System, and select the the datastore_fs log file. Find the time frame for when the VMs had export re-enabled, this should give clues as to why the vmx files are not being found by vmPro.
-
-
Jerry, each of the offending VMs is on a different datastore, but each is also fully contained on it's datastore. Still have the problem of the missing .vmx files. Our IT guy is thinking this may be a Vmware tools issue and is intending to update the tools on the VMs. But nothing has changed and things were working fine until the 29th. This is what I got from the datastore_fs log file:
2013-12-05 09:28:26.680612: datastore_fs_open(path /4da602d1-607dc2b6-e9e3-78e7d17d6278/rails/rails.vmx, fuse_file_info 0x7f4ea2304e40): direct_io 0, Random I/O
GET https://vcenter.corp.ad/folder/rails/rails.vmx?dsName=FataDataStore... unsuccessful : 500 Internal Server Error
2013-12-05 09:28:27.130931: ERROR [thread:139975705253632, where:/usr/local/pancetera/Perl/Pancetera/VIRefreshNS.pm:124, error:5] [FataDataStore] rails/rails.vmx: 500:Internal Server Error: Input/output error
datastore_fs[2865]: ERROR [thread:139975705253632, where:/usr/local/pancetera/Perl/Pancetera/VIRefreshNS.pm:124, error:5] [FataDataStore] rails/rails.vmx: 500:Internal Server Error: Input/output error
2013-12-05 09:28:27.349810: datastore_fs_open(path /4c9a55c6-389c6e72-82f4-0021286b6e4e/quoteshare/quoteshare.vmx, fuse_file_info 0x7f4ea0f02e40): direct_io 0, Random I/O
2013-12-05 09:28:27.404998: datastore_fs_open(path /50c43db0-d4531606-4a53-00237ddf1c4c/acct/acct.vmx, fuse_file_info 0x7f4ea2304e40): direct_io 0, Random I/O
GET https://vcenter.corp.ad/folder/quoteshare/quoteshare.vmx?dsName=EVA... unsuccessful : 500 Internal Server Error
-
-
Via vmPro do you connect to a vCenter or ESX server(s)?
You could try, disconnecting to the vCenter or ESX server(s) and then reconnect. If you are connected to a vCenter then try disconnecting to the vCenter and just connect to the ESX server(s) instead to see if that resolves the issue.
-
-
Jerry, thanks for you help on this. Over the weekend our IT guy identified some issues with one of the nodes in the VM cluster and issues with the vcenter server which caused it to disconnect. He worked on both of those issues and while it's not completely fixed yet we got good backups of all of the VMs over the weekend.
So for now it's working, and in he meantime we'll work on getting the cluster healthy again.
-
-
I'm having a similar issue. I'm using vmPro to backup my lab server. One ESX 5.1 host. A Windows 7 Pro management PC. Backup target is 700 GB shared via CIFS from the Win PC. Attempting to run a backup yields an instant "VM was not found in /export..." message for each VM. If I browse the vmPro VM from Windows, I see eight shares, including "export". The export share has the node folder in it, but that folder is empty. I'd really like to get this working, because it would give me a safety net for my lab activities other than VM snapshots. I have SSH running on the host, and when I toggle the "Exported" setting, I see the VM snapshots occurring in the vCenter client. Looks like I'm missing just one thing...
© 2024 Created by Quantum Forum V.
Powered by
