Networknet.nl Blog

Last night we lost our Openfiler SAN with a “kernel panic” stop message and all virtual machines located on the VMWare ESX cluster were frozen. After I found the problem and I reboot the SAN box;  I than encountered another problem on one of our development SQL server.  Just after the POST phase a black screen was shown and boot process stopped. I did not create any screenshot of this window but it was something like:

Windows  Server could not start because the following file is missing or corrupt: \WINDOWS\SYSTEM32\CONFIG\SYSTEM

Last time I had this error I was not able to fix this problem without reinstalling the OS. No recent backup of this VM and I run out of my options  to get this VM up and running again.

What I did than:

• Shutdown the VM
• Shutdown another working VM
• Mount the virtual disk of the failed Windows installation on the second VM and boot.
• After Windows was boot of the second vm I assigned a disk letter of the failed Windows instllation.

I went to E:\Windows\System32\config and got the following error message.

\Windows\System32\config is not accessible. The file or directory is corrupted and unreadable.

I than right clicked on the volume; went to properties and opened tools tab. I clicked “check now” to repair the volume. After 9% I got warning volume and the process stopped.  I was lost again. But than I remembered one tool which I tried than.

chkdsk e: /f

After the process and check & fix were finished I was able to access all folders. I than shutdown the VM disconnected the virtual disk and boot the development SQL server. After 5 minutes the machine was up and running. This saved me at least couple of days of work to restore the configuration and make everything working.

Windows showed me the warning ” Windows-Registery Hive Recovered”. I was again being able to logon on the server and make sure proper backup was in place now. Same can be done with physical disk but make sure you install the HDD into another machine!

Recently the motherboard of one of the HP ProLiant ML370 G4  was replaced due faulty hardware, but the ILO was never reconfigured back. The ILO board was not reachable anymore on the old IP address. Last weekend we had issues with Active Directory and none of the domain accounts were able to logon locally with a RDP session; so the ops team tried the ILO remote web interface and this failed. Today I was able to find the IP address of the ILO board with the help of the “HP System Management Home Page” tools. I tried to login with the known password but the authentication failed. Couple years ago when these HP servers were deployed there is option in the bios to reset the ILO setting and change the ip and username setting. I do not want to reboot the server and manually reconfigure the ILO board from the console. HP has command line tool for this named HPONCFG. Please go here or search for HPONCFG on google.

HP Lights-Out Online Configuration Utility (HPONCFG)
Version 1.5.2.0

Description
———–

HP Lights-Out Online Configuration utility is a command line utility used to configure
iLO, iLO 2 and RILOE II from within the Windows operating systems without requiring a
reboot of the server.

I am not able to show the screenshots of HPONCFG.exe tool. The server is managed by third party, but after you installed the cpqsetup.exe on the server you should be able to configure the ILO board.