Description of problem: RH9 kickstart/NFS installation fails with anaconda throwing "unhandled exception". Version-Release number of selected component (if applicable): RH9; anaconda-9.0-4 and anaconda-runtime-9.0-4. How reproducible: Attempt to start kickstart using Solaris NFS server as distribution source. The installer will ask for the IP number and default gateway for the machine to be installed. After (during?) the "Reading packages" phase, the unhandled exception is seen, and no further progress is made. Steps to Reproduce: 1.Boot custom RH9 cd built from RH9 cd 1/3 isolinux tree and containing custom isolinux.cfg and ks.cfg files. 2.Enter IP number and gateway for workstation. 3.Exception thrown after "Reading packages" is displayed. Actual results: Kickstart installation fails. Expected results: Installation would succeed. Additional info: Components involved: custom RH9 boot CD, custom isolinux.cfg and ks.cfg files in root of this boot CD, Solaris NFS server with RedHat9 CDs copied to an NFS shared filesystem, Addition of three packages to the RedHat distribution, with modification of comps.xml, and regeneration of hdlist and hdlist2 files.
It will help us identify the issue if the contents of the exception window were made available as an attachment to this report.
Created attachment 93305 [details] Anaconda dump for "unhandled exception" during RH9 kickstart When kickstart halts, VC2 (Alt+F2) shows that the NFS server has been properly mounted on /mnt/source. VC3 (Alt+F3) has the following as the last line of output: * Using file:///mnt/source/./RedHat/base/comps.xml I will try rebuilding this distribution without my added packages (i.e., with a "clean" comps.xml file) to see if this, rather than anaconda, is the source of the problem.
Great let us know how it works out.
Well, the comps.xml file appears to be the source of the problem. Building a clean distribution without additions allow kickstart to proceed smoothly. Consider this resolved.