The only error/output we get is:
2012-02-29 17:39:58 -- info: ERROR: Unable to locate *vimsh*! You're not running ESX(i) 3.5+, 4.x+ or 5.0!
We are running ESXi 5 and vMA 5
Enterprise build 515841
vMA build 472630
Backup datastore target is iSCSI, via OpenFiler
I posted a similar request in another thread a few days ago (in which someone posted having a very similar problem description). However, I realized afterwards that it was in a vsphere 4 discussion group, so I thought this would be a more appropriate place to post.
We have used VMA + GhettoVCB as part of our backup strategy for several years. Recently our engineer implemented vCenter 5, upgraded our hosts to vSphere 5, and deployed the latest VMA appliance.
Our old setup was using the .pl scripts and worked fine. With the new stuff, we are running in to a wall, it seems. We supposed that we still needed to configure fastpass, and when we do vifp listservers we see our hosts, so I think we're good there.
However, when we attempt to run the ghettoVCB.sh script, we get this:
"ERROR: Unable to locate *vimsh*! You're not running ESX(i) 3.5+, 4.x+ or 5.0!"
Of course we are running ESXi 5. Searching Google and even these forums for the error is very difficult, because the error is part of the script. So the search results return copy/paste versions of the script itself, not people having this issue. So far, this forum post from July 2010 is the closest I could find to anyone having the same problem.
Looking at the code of the script, it appears to be looking for some particular folders. When those folders aren't found, it tells you that you're not running ESX. I can't tell if it is looking for these folders on the VMA itself or on the vSphere hosts. My thoughts at this point is that we have some permissions issue or perhaps some issue with vSphere 5 having a software firewall in place. I'm kind of doubting it's a v5 issue because bklynjae appears to have had the same problem on a v4 host.
If we can get 1 to work, I'm sure we can get the rest to work. So, we put 1 VM on a host by itself, and we tried with and without putting in the vifptarget command to the host we wanted to backup. We filled out the ghettoVCB.conf file and then we tried this command:
./ghettoVCB.sh -a -d dryrun
He also tried variations attempting to use our ghetto.conf file, vmlist, etc. Regardless, we are greeted with the error message regarding vimsh. My engineer is sure that it will end up being something trivial and only require seconds to resolve... once we know what to change...
Thanks in advance for any assistance.