Re: panic, but /var/crash ist empty
Brandon Gooch wrote: > On Wed, Apr 27, 2011 at 1:44 PM, Helmut Schneider > wrote: > > What can I do to create a backtrace to open a PR? > > To get a backtrace from the crash (or drop to the debugger), you'll > need to compile a kernel with at least a couple of options defined. > These two: > > options KDB > options DDB > > ...will allow you to work with the debugger on the console after a > crash. > > Further, with the option KDB_TRACE in the kernel config, you'll get a > backtrace printed automatically when the kernel panicsb. > > Here are a couple of excellent documents to read to get you started: > > http://www.freebsd.org/doc/en/books/handbook/kernelconfig-building.html > http://www.freebsd.org/doc/en/books/developers-handbook/kerneldebug.html > > Kernel debug options: > > http://www.freebsd.org/doc/en/books/developers-handbook/kerneldebug-options.html Thanks for the links, I meanwhile was able to create the bt and the PR: http://www.freebsd.org/cgi/query-pr.cgi?pr=156691 Helmut ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: panic, but /var/crash ist empty
On Wed, Apr 27, 2011 at 1:44 PM, Helmut Schneider wrote: > Hi, > > running 8.2-RELEASE-p1 within VMWare ESXi 4.1-u1 I want to use raw > devices as hard disks. I create the devices using this link: > > http://www.mattiasholm.com/node/33 > > I tried 3 different hard drives (Seagate 2x80GB and 1x400GB SATA2) > which are fine on a physical machine. I also ran Seatool many hours on > all of them without errors. > > I can partiton the disks and create a few files/directories on it. But > as soon as I copy a larger number of files to those disks (tried with > MBR and GPT) the VM reboots *instantly* (I tried cp, dump/restore and > rsync). No "Rebooting within 15 seconds", just *snap*. I think I can > see an panic but I'm not sure, it's too fast. > > (as far as I can see most of the times the data on the first UFS slice > (and only the first UFS slice!) of the partition gets *severly* > corrupted, most of the time all that is left are a few files within > lost+found. Sometimes all the labels are gone but are recoverable using > bsdlabel -R) > > The problem is that /var/crash remains empty. > > What can I do to create a backtrace to open a PR? > > Thanks, Helmut > Hi Helmut, To get a backtrace from the crash (or drop to the debugger), you'll need to compile a kernel with at least a couple of options defined. These two: options KDB options DDB ...will allow you to work with the debugger on the console after a crash. Further, with the option KDB_TRACE in the kernel config, you'll get a backtrace printed automatically when the kernel panicsb. Here are a couple of excellent documents to read to get you started: http://www.freebsd.org/doc/en/books/handbook/kernelconfig-building.html http://www.freebsd.org/doc/en/books/developers-handbook/kerneldebug.html Kernel debug options: http://www.freebsd.org/doc/en/books/developers-handbook/kerneldebug-options.html ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
Re: panic, but /var/crash ist empty
Helmut 'Ingrid' Schneider wrote: > running 8.2-RELEASE-p1 within VMWare ESXi 4.1-u1 I want to use raw > devices as hard disks. I create the devices using this link: > > http://www.mattiasholm.com/node/33 > > I tried 3 different hard drives (Seagate 2x80GB and 1x400GB SATA2) > which are fine on a physical machine. I also ran Seatool many hours on > all of them without errors. > > I can partiton the disks and create a few files/directories on it. But > as soon as I copy a larger number of files to those disks (tried with > MBR and GPT) the VM reboots instantly (I tried cp, dump/restore and > rsync). No "Rebooting within 15 seconds", just snap. I think I can > see an panic but I'm not sure, it's too fast. > > (as far as I can see most of the times the data on the first UFS slice > (and only the first UFS slice!) of the partition gets severly > corrupted, most of the time all that is left are a few files within > lost+found. Sometimes all the labels are gone but are recoverable > using bsdlabel -R) > > The problem is that /var/crash remains empty. > > What can I do to create a backtrace to open a PR? FWIW: While a sysinstall from CD (tried 8.0, 8.1 and 8.2) also dies short before the end of the installation, Ubuntu 10.4 (EXT4), Windows 7 and OpenBSD install fine and also do not die when copying a lots of files... :( ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
panic, but /var/crash ist empty
Hi, running 8.2-RELEASE-p1 within VMWare ESXi 4.1-u1 I want to use raw devices as hard disks. I create the devices using this link: http://www.mattiasholm.com/node/33 I tried 3 different hard drives (Seagate 2x80GB and 1x400GB SATA2) which are fine on a physical machine. I also ran Seatool many hours on all of them without errors. I can partiton the disks and create a few files/directories on it. But as soon as I copy a larger number of files to those disks (tried with MBR and GPT) the VM reboots *instantly* (I tried cp, dump/restore and rsync). No "Rebooting within 15 seconds", just *snap*. I think I can see an panic but I'm not sure, it's too fast. (as far as I can see most of the times the data on the first UFS slice (and only the first UFS slice!) of the partition gets *severly* corrupted, most of the time all that is left are a few files within lost+found. Sometimes all the labels are gone but are recoverable using bsdlabel -R) The problem is that /var/crash remains empty. What can I do to create a backtrace to open a PR? Thanks, Helmut ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"
