The current version of the Network Preparation Script expects a full Class C subnet mask in the NIC used for GUI/API access (ifcfg-eth0 or ifcfg-eth1).

If you change the subnet mask during the execution of the script, the script will replace your subnet mask (for example, if you enter 255.255.248.0, the script will replace the subnet mask with 255.255.255.0).

This issue will be addressed in the next release of the Network Preparation Script. In the meantime, the workaround is, after the initial script failure:

1) Edit /etc/sysconfig/network-scripts/ifcfg-eth0

2) Change NETMASK=255.255.255.0 to NETMASK=255.255.248.0

2.1) if the NETMASK line does not exist append it as NETMASK=255.255.248.0

3) Validate with a network restart

# service network restart

4) Execute the script again

sh cisco-prep-network-{version_number}.exp

This should show the 255.255.248.0 as the discovered default. Accept the default and continue.

If unable to run the Network Preparation Script to completion, collect the *status* and *log* files:

# cd /usr/local/linmin

# zip {YourCompanyName}-prepServerLogs.zip *_status* *log*

If you are unable to run the Server Preparation Script to completion, exit the script with Ctrl+C, then, when in /usr/local/linmin:

# ./cisco--prep-server-{version_number}.exp

When prompted, select the default "extract":

Respond: 'extract' the package contents or 'reuse' the previously extracted modules:extract:reuse:quit: [extract] :

Select the option to create a help file:

Quit or create a help file

^^^^^

Respond: ^^^^^ 'q' to exit or 'c' to create help file :c:q: [q]: c

Then answer the questions to properly label the .help file. The result will be similar to:

Created help file: bmp_Brian_Jones_ATT_linmin-prep-server_20110816-090259.help

-rw-r--r-- 1 root root 8.5K Aug 16 09:04 bmp_Brian_Jones_ATT_linmin-prep-server_20110816-090259.help

Email the help file to Support: bmp_Brian_Jones_ATT_linmin-prep-server_20110816-090259.help

Ignore messages stating that the script was interrupted (it was interrupted since you were unable to prepare your system).

Email the file(s) to Support with as many details of your experiences as possible: system you are running on (physical or virtual), OS (RHEL or CentOS version), host hypervisor, etc.

Note: .zip files may be stripped by mail servers. Contact Support to request FTP site credentials to upload your file(s).

Downloaded files and packages can sometimes be corrupted. Provisioner PXE Server packages contain built-in MD5 verification. Should you encounter such an error, download the package again and retry installing it.

Below is an example of such an MD5 sum error due to a corrupted download, encountered after attempting to install the package:

...locate start of archive
 
...checking package archive md5
 
... This may take several minutes. Please do not interrupt...
 
66add5fb626785278feed29d62006d00 : packaged archive md5
 
d52ff5aa8b369930c5731f975436d42c : package archive calculated md5
 
^^^^^^^^^^ calculated package archive MD5 not equal packaged archive MD5
 
!!! package is corrupt
 
Respond: ^^^^^ read above, and reply 'a' to acknowledge  :a: [] :a

Presently, installation of Provisioner 6.3 on RHEL 6.x or CentOS 6.x is not supported.

Should you attempt to install Provisioner 6.3 on RHEL 6.x or CentOS 6.x, you will get an error message similar to the one below:

... execute the install script

cisco-prep-server-6.3.0-12-p8.exp: ./linmin-prep-server.sh: /lib/ld-linux.so.2: bad ELF interpreter: No such file or directory

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

•Did you hit Control-C (thinking the installation was taking too long) or did the system simply hang? If so, you may be able to uninstall and re-install, however if this fails, you may need to re-install your OS and start the server preparation and installation over.

•Do you have Internet access or or access to a software repository (e.g., Red Hat Network access?) Do a yum update to verify.

•Is your system clock set correctly? The Quick Start Guide requires that you set NTP to have a correct time on your system. If you failed to do so, you may be able to uninstall and re-install, however if this fails, you may need to re-install your OS and start over (Linux does not gracefully handle working with files dated from the future.)

•Is your firewall turned off or do you have the correct firewall ports open?

•Are you running VMware with a newly cloned VM of RHEL or CentOS? After cloning a VM, power it up then reboot it before installing the Provisioner PXE Server.

•Did you run out of disk space (on a physical box, or with VMware, on either the guest or host OS?)

If you are encountering network problems, please verify that the corresponding parameters’ values in these files match:

/home/tftpboot/config/ocprovision.conf

/etc/dhcpd.conf

/etc/syconfig/network-scripts/ifcfg-eth0

/etc/resolv.conf.

Edit and correct above files. After making any changes, execute the following:

/usr/local/linmin/status/linmin-services.sh restart

Sometimes it's best to simply uninstall and re-install. After you uninstall, you will need to recreate /usr/local/linmin/ and copy the Provisioner PXE Server files to this directory. If you use VMware, revert to an earlier snapshot and try again.

If you still require assistance, please follow the instructions in "Contacting Technical Support".