[user@box rpm]$ sudo rpm -ivh --force veeamsnap-1.0.0.678-1.noarch.rpm
Preparing... ########################################### [100%]
1:veeamsnap ########################################### [100%]
Removing old veeamsnap-1.0.0.678 DKMS files...
------------------------------
Deleting module version: 1.0.0.678
completely from the DKMS tree.
------------------------------
Done.
Loading new veeamsnap-1.0.0.678 DKMS files...
Building for 2.6.32-573.12.1.el6.x86_64
Module build for kernel 2.6.32-573.12.1.el6.x86_64 was skipped since the
kernel headers for this kernel does not seem to be installed.
WARNING: Package not configured! See output!
warning: %post(veeamsnap-1.0.0.678-1.noarch) scriptlet failed, exit status 1
[user@box rpm]$ rpm -qa | grep kernel-headers
kernel-headers-2.6.32-573.12.1.el6.x86_64
[user@box rpm]$ uname -r
2.6.32-573.12.1.el6.x86_64
So, the only entry in /usr/src/kernels is for a different revision, 2.6.32-642.4.2.el6.x86_64. But it is not a symlink. /lib/modules has a folder for the correct version.
It might be related to the headers not loaded. This happens in rare occasions and a reboot should resolve it.
However I have seen other things as well which require the following:
- remove veeam packages
- remove headers
- install headers
- reboot
- install veeam packages
I don't know if you can reboot the server but if possible please try the above and see if it works.
Please try to erase everything including dkms. Afterwards install dkms from scratch, cd to the directory where veeamsnap rpm is and try install it via yum install ./veeamsnap*, that should let yum to pull and install all dependencies automatically.