From Bugzilla Helper: User-Agent: Mozilla/4.76 [en] (X11; U; Linux 2.4.2-0.1.49smp i686) When the up2date was done doing its rpm -ivh's, a window poped up saying that the test version of lilo failed Reproducible: Didn't try Steps to Reproduce: 1.run up2date 2.request that kernel gets updated 3.watch error message window pop up The problem is that there was no output or log or anything which I could determine why the test lilo failed. My lilo file, which I'll attach is *very* generic. There should have been no problem with it. One bit which is strange on my system, and I don't know if I should bugzilla this is that I have two entries for the same rpm package, specifically the kernel rpms. I caused this in my rash of rpm'ing one day and I installed the kernel rpm twice with a --force. I can't seem to get rid of it now. Here look: [adler@ssadler adler]$ rpm -qa | grep kernel kernel-headers-2.4.2-0.1.49 kernel-source-2.4.2-0.1.49 kernelcfg-0.6-10 kernel-doc-2.4.2-0.1.49 kernel-pcmcia-cs-3.1.24-2 kernel-2.4.2-0.1.28 kernel-enterprise-2.4.2-0.1.28 kernel-BOOT-2.4.2-0.1.49 kernel-smp-2.4.2-0.1.49 <<<<----????? kernel-smp-2.4.2-0.1.49 <<<<----????? kernel-2.4.2-0.1.49 <<<<----????? kernel-2.4.2-0.1.49 <<<<----????? kernel-BOOT-2.4.2-0.1.28 kernel-enterprise-2.4.2-0.1.49 kernel-smp-2.4.2-0.1.28 What I've done is modified the lilo.conf file by hand, I've run lilo and I'm now runnig the new kernel.
Created attachment 14804 [details] original lilo.conf file which up2date saw when up2dating the kernel
multiple versions of kernels packages is expected. Multiple instances of the same kernel package (ie, the two kernel-smp-2.4.2-0.1.49 you show) is not. In theory, thats not supposed to happen ;-> Of course, in practice, it does happen on occasion. More than likely one is a 386 kernel, and one is a 686 kernel. Can you try: rpm -qa --queryformat "%{NAME}-%{VERSION}-%{RELEASE}.%{ARCH}\n" | grep kernel I'll attempt to duplicate this. I'm assuming that up2date gave the error message about lilo failing, but left the system in a bootable state? Making sure we successfully install the new kernels is important, but making sure we handle failures gracefully in this case is probabaly even more important.
The output you requested. Looks like I do have multple architectures installed. How does one remove the unwanted architecture? Steve. [root@ssadler /root]# rpm -qa --queryformat "%{NAME}-%{VERSION}-%{RELEASE}.%{ARCH}\n" | grep kernel kernel-headers-2.4.2-0.1.49.i386 kernel-source-2.4.2-0.1.49.i386 kernelcfg-0.6-10.i386 kernel-doc-2.4.2-0.1.49.i386 kernel-pcmcia-cs-3.1.24-2.i386 kernel-2.4.2-0.1.28.i686 kernel-enterprise-2.4.2-0.1.28.i686 kernel-BOOT-2.4.2-0.1.49.i386 kernel-smp-2.4.2-0.1.49.i686 kernel-smp-2.4.2-0.1.49.i586 kernel-2.4.2-0.1.49.i386 kernel-2.4.2-0.1.49.i686 kernel-BOOT-2.4.2-0.1.28.i386 kernel-enterprise-2.4.2-0.1.49.i686 kernel-smp-2.4.2-0.1.28.i686
I belive all the bugs related to multiple kernels should be well shaken out at this point, so closing this bug.