ghettoVCB: Backups now failing when using VMDK_FILES_TO_BACKUP
It looks like recently PR #198 was merged in. I wonder how throughly this was tested, as I now have multiple VMs failing that used to backup just fine before.
Here is a snippet from the log:
...
2021-01-08 00:30:12 -- info: CONFIG - VMDK_FILES_TO_BACKUP = Disk-0.vmdk
...
2021-01-08 00:30:15 -- debug:
2021-01-08 00:30:15 -- debug: getVMDKs() - Disk-2.vmdk###250:Disk-0.vmdk###32:
2021-01-08 00:30:15 -- info: WARNING: Disk-0.vmdk not found in VMDKs for MonoVM
2021-01-08 00:30:15 -- info: ERROR: No valid VMDKs in list of VMDKs to backup for MonoVM
2021-01-08 00:30:15 -- info: ERROR: Failed to backup MonoVM!
2021-01-08 00:30:15 -- debug: VM_OK:0, VM_FAILED:1, VMDK_FAILED:0
2021-01-08 00:30:15 -- info: ###### Final status: ERROR: All VMs failed! ######
It looks like the call to getVMDKs returns a colon separated string:
https://github.com/lamw/ghettoVCB/blob/42b119f9dad9f08109c581de6efe9e785e91287a/ghettoVCB.sh#L500
But in the new function to iterate over this to check, there is no code that attempts to loop over this colon, but rather it is looking for a “,”:
I am not sure how this is supposed to work at all.
About this issue
- Original URL
- State: closed
- Created 3 years ago
- Comments: 15 (5 by maintainers)
Can verify it works. Had the same issue this week after rebuilding our esxi. Backup only worked when just one vmdk was present in the $VMDKS as the list wasn’t properly separated.
Tested on two VMs.
It worked with your suggestion. I also tried it with the system generated symlink (for the datastore name) and that did not work. Is there going to be any fix for using the symlink or not having to use the full path? Otherwise, I’m sure to forget to change this when the VMDKs get moved.
thx for your help
On Sat, Aug 7, 2021 at 2:40 AM DAN @.***> wrote: