Help to find cause of recurring crash

[Pietro Cerutti]

On 5/16/06, fbsd <fbsd at aanet.com.au> wrote:
> Pietro Cerutti wrote:
> > On 5/16/06, fbsd <fbsd at aanet.com.au> wrote:
> >> Hello All,
> >>
> >> I'm  running a home email server and twice in the last week it has
> >> rebooted itself and been unable to restart due to corruption of the
> >> filesystem. I found the following in /var/log/messages
> >>
> >> May 16 17:46:07 hpvectra syslogd: kernel boot file is
> >> /boot/kernel/kernel
> >> May 16 17:46:07 hpvectra kernel:
> >> May 16 17:46:07 hpvectra kernel:
> >> May 16 17:46:07 hpvectra kernel: Fatal trap 12: page fault while in
> >> kernel mode
> >> May 16 17:46:07 hpvectra kernel: fault virtual address    = 0x1c
> >> May 16 17:46:07 hpvectra kernel: fault code        = supervisor write,
> >> page not present
> >> May 16 17:46:07 hpvectra kernel: instruction pointer    =
> >> 0x20:0xc062c5e8
> >> May 16 17:46:07 hpvectra kernel: stack pointer            =
> >> 0x28:0xe5079c50
> >> May 16 17:46:07 hpvectra kernel: frame pointer            =
> >> 0x28:0xe5079c64
> >> May 16 17:46:07 hpvectra kernel: code segment        = base 0x0, limit
> >> 0xfffff, type 0x1b
> >> May 16 17:46:07 hpvectra kernel: = DPL 0, pres 1, def32 1, gran 1
> >> May 16 17:46:07 hpvectra kernel: processor eflags    = interrupt
> >> enabled, resume, IOPL = 0
> >> May 16 17:46:07 hpvectra kernel: current process        = 52 (vnlru)
> >> May 16 17:46:07 hpvectra kernel: trap number        = 12
> >> May 16 17:46:07 hpvectra kernel: panic: page fault
> >> May 16 17:46:07 hpvectra kernel: Uptime: 5d7h4m50s
> >> May 16 17:46:07 hpvectra kernel: Dumping 1023 MB (2 chunks)
> >> May 16 17:46:07 hpvectra kernel: chunk 0: 1MB (159 pages) ... ok
> >> May 16 17:46:07 hpvectra kernel: chunk 1: 1023MB (261872 pages) 1007 991
> >> 975 959 943 927 911 895 879 863 847 831 815 799 783 767 751 735 719 703
> >> 687 671 655 639 623 607 591 575 559 543 527 511 495 479 463 447 431 415
> >> 399 383 367 351 335 319 303 287 271 255 239 223 207 191 175 159 143 127
> >> 111 95 79 63 47 31 15 ... ok
> >> May 16 17:46:07 hpvectra kernel:
> >> May 16 17:46:07 hpvectra kernel: Dump complete
> >> May 16 17:46:07 hpvectra kernel: Automatic reboot in 15 seconds - press
> >> a key on the console to abort
> >> May 16 17:46:07 hpvectra kernel: Rebooting...
> >
> > The only thing I can see (read: understand) is:
> > the process that caused the kernel panic is "vnlru":
> >
> > "vnlru flushes and frees vnodes when the system hits the
> > kern.maxvnodes limit. This kernel thread sits mostly idle, and only
> > activates if you have a huge amount of RAM and are accessing tens of
> > thousands of tiny files."
> >
> > (from http://www.unixguide.net/freebsd/faq/10.31.shtml)
> >
> > So, consider increasing kern.maxvnodes if your system deals with such
> > a huge amount of files.
> >
> >>
> >
> >> Thanks in advance,
> >
> > Hope this helps,
> >
> >>
> >> Ron
> >
> >
> Thanks Pietro,
>
> I will do that. However I don't think 1GB of RAM is so big and as a home
> server I would be lucky to deal with more than a couple of hundred
> emails per day.

I wouldn't trust in suppositions ;-)

check whether vfs.maxvnodes (max vnodes allowed) minus vfs.numvnodes
(number of actual vnodes) is nearly zero.

>So if anyone can shed any more light it will be appreaciated.
>
> Thanks
>
> Ron
>

-- 
Pietro Cerutti
icq #117293691

- ASCII Ribbon Campaign -
 against HTML e-mail and
 proprietary attachments
   www.asciiribbon.org