|Red Hat Enterprise Linux 4: Installation Guide for the IBM® POWER Architecture
|Appendix B. Troubleshooting Your Installation of Red Hat Enterprise Linux
If you receive an error message stating No devices found to install Red Hat Enterprise Linux, there is probably a SCSI controller that is not being recognized by the installation program.
Check your hardware vendor's website to determine if a driver diskette image is available that fixes your problem. Second, add dd to the IPL parameters line in the Network Storage Description. For more general information on driver diskettes, refer to Appendix D Driver Media.
You can also refer to the Red Hat Hardware Compatibility List, available online at:
If you receive a traceback error message during installation, you can usually save it to a diskette.
If you do not have a diskette drive available in your system, you can scp the error message to a remote system.
This information does not apply to users of iSeries systems or headless pSeries systems.
When the traceback dialog appears, the traceback error message is
automatically written to a file named
/tmp/anacdump.txt. Once the dialog appears, switch
over to a new tty (virtual console) by pressing the keys
If you receive an error after the Disk Partitioning Setup (Section 2.12 Disk Partitioning Setup) phase of the installation saying something similar to
The partition table on device hda was unreadable. To create new partitions it must be initialized, causing the loss of ALL DATA on this drive.
you may not have a partition table on that drive or the partition table on the drive may not be recognizable by the partitioning software used in the installation program.
No matter what type of installation you are performing, backups of the existing data on your systems should always be made.
If you are using Disk Druid to create partitions, but cannot move to the next screen, you probably have not created all the partitions necessary for Disk Druid's dependencies to be satisfied.
You must have the following partitions as a bare minimum:
A / (root) partition
A <swap> partition of type swap
A PPC PReP Boot partition.
For non-iSeries users only, a /boot/ partition.
When defining a partition's type as swap, you do not have to assign it a mount point. Disk Druid automatically assigns the mount point for you.
During some upgrades or installations of Red Hat Enterprise Linux, the installation program (also known as anaconda) may fail with a Python or traceback error. This error may occur after the selection of individual packages or while trying to save the upgrade log in the /tmp/directory. The error may look similar to:
Traceback (innermost last): File "/var/tmp/anaconda-7.1//usr/lib/anaconda/iw/progress_gui.py", line 20, in run rc = self.todo.doInstall () File "/var/tmp/anaconda-7.1//usr/lib/anaconda/todo.py", line 1468, in doInstall self.fstab.savePartitions () File "fstab.py", line 221, in savePartitions sys.exit(0) SystemExit: 0 Local variables in innermost frame: self: <fstab.GuiFstab instance at 8446fe0> sys: <module 'sys' (built-in)> ToDo object: (itodo ToDo p1 (dp2 S'method' p3 (iimage CdromInstallMethod p4 (dp5 S'progressWindow' p6 <failed>
This error occurs in some systems where links to /tmp/ are symbolic to other locations or have been changed since creation. These symbolic or changed links are invalid during the installation process, so the installation program cannot write information and fails.
If you experience such an error, first try to download any available errata for anaconda. Errata can be found at:
The anaconda website may also be a useful reference and can be found online at:
You can also search for bug reports related to this problem. To search Red Hat's bug tracking system, go to:
Finally, if you are still facing problems related to this error, register your product and contact our support team. To register your product, go to: