I was able to get VCB working at one point with VC 2.0 and EMC Notworker 7.3.x, but with VC 2.5 and Notworker 7.4.2, I cannot. I consistently receive the following errors:
The virtual machine identified by is unknown.
Exit Code: 16
So far, EMC support has been less than helpful, suggesting that I need to call VMWare. Has anyone else seen this problem?
Yes, make a backup of config.js, uninstall VCB and delete the integration module. Install VCB, unzip and install the integration module. Edit the settings in the NEW config.js to point to your mount point user and password.
If you are running VC2.5U2 and VCB 1.5, you can also use a user name of "_sspi_". The directions for setting this up are also in the guide. Basically, the legato service account needs to be a user in VC...
Dave
Try a query with the vcbVmName utility to make sure the name didn't actually change. Also, you can set LOGFILE=6 in config.js to get a verbose log. I would also upgrade to VCB 1.5 which was just released last Friday.
Dave
The vcbVmName command indicates that the VM is named what I think it's named. Where would the logfile go and what might it be named?
They will be located in %TEMP% and are named vmware-vmount.log and vmware-vlun.log. %TEMP% is usually C:\WINDOWS\TEMP unless you changed it.
Dave
I'm feeling the same pain bud,
Like you I used the VCB command to get the names to be accurate about the name (case and spelling etc) but it didn't help. The other method I try was using the IP address of the VM instead of the name for the client which I would have thought worked but it didn't.
The verbose logging for me was displayed only in the command prompt from where I executed the VCB mount command, that revealed that VCB didn't have a problem but Legato always returned 'what vm' and I can't think what else to try. So it must be something with Legato Networker, I was going to try and log a call with them instead.
You can try running the vcbMounter command directly to rule out VCB.
Dave
VcbMounter works, although I note what may be a problem: using name based access, the VM name is upper case, but Networker forces its client names to lower case, and the command is case-sensitive. If I use the ipaddr search directive with the VM hostname, that works fine, so it's clearly possible to mount the VM.
OK, now you're getting somewhere. Yes, everything is case sensitive with VCB. You should be able to use hard IP addresses or FQDNs to perform your backups. You may be able to set your config.js with VM_LOOKUP_METHOD="ipaddr";
I've never used Networker with VCB, but the underlying scripts are the same for all.
Dave
Right--as far as I can tell, VCB is configured correctly to use the ipaddr directive, which is why I'm thinking it's a matter of how the Networker module is talking to VMWare.
Maybe there is a glue.js that calls the VM name instead of the ipaddr??
Has EMC gotten back to you at all? I've been working with them to try to sort this out, but with no luck so far. The only thing I can think of is that VMWare is now using an SSL certificate in VC 2.5, which may make Networker unhappy.
One other thing...you mentioned that you upgraded. In config.js, change the port to 443. 902 does not work now.
Well, that was certainly part of it. Thank you very much for the info! I've moved on to the following output:
savegrp: suppressed 55 lines of output.
A certificate in the host's chain is based on an untrusted root.
SSLVerifyIsEnabled: failed to read registry value. Assuming verification is disabled. LastError = 0
SSLVerifyCertAgainstSystemStore: Certificate verification is disabled, so connection will proceed despite the error
Error: Other error encountered: vim.fault.NoPermission
An error occurred, cleaning up...
Deleted directory D:\vcb\mnt\
External command failed. See error above.
Exit Code: 1
I'm going to guess that this is related to permissions for the VCB service account, which currently has the VCB Proxy User role assigned to it for the entire data center. Are there other permissions I should be looking at?
I answered this question myself and changed the service account from the VCB Proxy User role to VMware Consolidated Backup User role, and that seems to be working. I am waiting to see how this turns out, but I'm feeling very optimistic!
Newest error is:
Version mismatch:
Exit Code: 10
Guessing that the solution is to remove VCB and/or the Notworker integration module and reinstall.
Yes, make a backup of config.js, uninstall VCB and delete the integration module. Install VCB, unzip and install the integration module. Edit the settings in the NEW config.js to point to your mount point user and password.
If you are running VC2.5U2 and VCB 1.5, you can also use a user name of "_sspi_". The directions for setting this up are also in the guide. Basically, the legato service account needs to be a user in VC...
Dave
Also, make sure you get the proper legato module for your version of ESX...
This appears to have done the trick. Thanks for all the good information! I am continuing to receive errors with some virtual machines but not all. The error codes are:
Error: Error while opening disk vpxa-nfc://:902!52 b0 51 2d de 0b d9 40-f9 0c 65 c7 6c 2c 6f 23: A virtual disk could not be opened. -- An invalid field name was used in a query string (Exception from HRESULT: 0x80040204)
and the ever-entertaining
Unknown error:
Exit Code: -2146234327
I'm going to create a separate thread for these errors, though.
Thanks again,
Tom
If you are running Virtual Center 2.5 / ESX 3.5, you should be using port 443 instead of 902. That may get rid of your errors..
Dave