Should the Provisioner PXE Server installation fail and result in a corrupted environment, it is generally attributed to:

1)The user interrupting the installation (for example, with a “Ctrl+C”) (90% of the time)

2)Hardware or network failure

3)“Yum update” not run prior to installing the Provisioner PXE Server (software repositories can’t be not found). As of release 6.0.2, the installation doesn't begin if no repository can be found due to lack of Internet connection or lack of a valid Red Hat Network subscription

4)Wrong RHEL/CentOS language selected (must be US English)

The recommended approach to recover from a failed installation is to use the supplied uninstall script. This handles the numerous extended configurations to the companion applications like Postgres. Below is the command stream for replacing a corrupt environment. Ignore any number of exception messages that might be presented in the uninstall.

To uninstall the Provisioner PXE Server, perform the following steps:

Execute as root user:

cd /usr/local/linmin

 

./lbmp-uninstall.sh

 

This moves /usr/local/linmin/ to /usr/local/linmin_{uninstall_date}/ and deletes /usr/local/linmin/.

 

To re-install the product, simply recreate the linmin/ directory, and retrieve the files that were moved to the uninstall directory. For example, to retrieve the base product installation package:

cd /usr/local/

mkdir linmin

cd linmin_{uninstall_date}

mv cisco-provisioner-{version}.exp ../linmin

cd ../linmin

 

And proceed to install the package(s).