On Monday 13 October 2008 09:31:50 Frédéric BOITEUX wrote: > Le dim 12 oct 2008 20:39:13 CEST, Joseph Rawson <email@example.com> a > > écrit : > > I have been using the net images a lot, and I have been experiencing > > intermittent problems when the ipconfig is being run. The init process > > always pauses at this point, often for a minute or so. I've noticed that > > if it pauses for longer than a couple of minutes, it will stay paused > > there forever. > > > > Using the debug option to the kernel makes matters even worse. While it > > puts more info into the live.log file, while booting, the debug > > information is hidden from the output, making debug less informative than > > having it off. In the circumstances where the init process stalls, the > > debug option is entirely useless (unless there's a way to invoke a shell > > at that point. I don't know how to do that). > > > > I'd like to be able to fix this problem. Can anybody tell me what to do > > about it? > > About the debug option, make sure you're using latest live-helper and > live-initramfs packages : the 'debug' behaviour has changed, and now > debug output is also sent to stdout (I use debug=whatyouwant as kernel > command line parameter, but I think 'debug' lonely should be enough). > I'll try a new one soon. > About ipconfig, I'm experiencing the same problem : if it fails, you > can't do anything but reboot... It does make me feel a little better that I'm not the only one experiencing this problem. It lets me know that it's probably not my configuration. Even so, I'd like to figure out why this is the only place that it seems to stall. I did some digging around to figure out where ipconfig comes from. I originally thought that it came from busybox, but it seems to belong to klibc-utils. Also, it seems that there's already a bug that has been filed against it - http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=432977 . Reading the bug reports, it seems that some people were having a hard time reproducing this, and it was closed and assumed to be fixed in March 2008. The version that was supposed to fix this was 1.5.8 and I'm using 1.5.12 . I'm going to try and dig a little further, and I'll let you know what I find. -- Thanks: Joseph Rawson
Description: This is a digitally signed message part.