Re: machine falling into a non-usable state for no apparent reason
Am Mittwoch, 29. Dezember 2004 17:47 schrieb Thomas McLean:
> Hi all,
>
> I'm running SID on 2.6.9 (no other kernel was installed before the
> current one). I've got a problem where my box randomly hangs and
> doesn't allow any connections in or out - ssh, telnet, finger, httpd -
> nothing except icmp ping requests. I've ran memtest and that seems to
> log the X session I'm currently in out. I also get the following error
> messages after it terminates itself from /var/log/messages:
>
> Dec 29 15:51:02 fresh kernel: memtest: page allocation failure.
> order:0, mode:0xd2
> Dec 29 15:51:02 fresh kernel: [<c0132b5e>] __alloc_pages+0x29f/0x2bc
> Dec 29 15:51:02 fresh kernel: [<c013b09a>] do_anonymous_page+0x63/0x14e
> Dec 29 15:51:02 fresh kernel: [<c013b1d4>] do_no_page+0x4f/0x2cf
> Dec 29 15:51:02 fresh kernel: [<c013b588>] handle_mm_fault+0x6c/0x125
> Dec 29 15:51:02 fresh kernel: [<c013a366>] get_user_pages+0x194/0x306
> Dec 29 15:51:02 fresh kernel: [<c013b6d6>] make_pages_present+0x6b/0x85
> Dec 29 15:51:02 fresh kernel: [<c013ba62>] mlock_fixup+0x76/0x8c
> Dec 29 15:51:02 fresh kernel: [<c013bacc>] do_mlock+0x54/0x99
> Dec 29 15:51:02 fresh kernel: [<c013bba2>] sys_mlock+0x91/0xb1
> Dec 29 15:51:02 fresh kernel: [<c0106073>] syscall_call+0x7/0xb
> Dec 29 15:51:02 fresh kernel: memtest: page allocation failure.
> order:0, mode:0xd2
> Dec 29 15:51:02 fresh kernel: [<c0132b5e>] __alloc_pages+0x29f/0x2bc
> Dec 29 15:51:02 fresh kernel: [<c01405ea>] read_swap_cache_async+0x54/0xba
> Dec 29 15:51:02 fresh kernel: [<c013adca>] swapin_readahead+0x3d/0x80
> Dec 29 15:51:02 fresh kernel: [<c013ae6d>] do_swap_page+0x60/0x22a
> Dec 29 15:51:02 fresh kernel: [<c013b5ae>] handle_mm_fault+0x92/0x125
> Dec 29 15:51:02 fresh kernel: [<c01150fe>] do_page_fault+0x19a/0x4ec
> Dec 29 15:51:02 fresh kernel: [<c01196cc>] call_console_drivers+0xe3/0xeb
> Dec 29 15:51:02 fresh kernel: [<c0119949>] release_console_sem+0x4a/0xc0
> Dec 29 15:51:02 fresh kernel: [<c011989b>] vprintk+0xfa/0x11a
> Dec 29 15:51:02 fresh kernel: [<c011666f>] scheduler_tick+0x317/0x3df
> Dec 29 15:51:02 fresh kernel: [<c0106d68>] print_context_stack+0x18/0x52
> Dec 29 15:51:02 fresh kernel: [<c0114f64>] do_page_fault+0x0/0x4ec
> Dec 29 15:51:02 fresh kernel: [<c0106af5>] error_code+0x2d/0x38
> Dec 29 15:51:02 fresh kernel: [<c019dea0>] __copy_to_user_ll+0x196/0x1b6
> Dec 29 15:51:02 fresh kernel: [<c011d260>] __do_softirq+0x34/0x79
> Dec 29 15:51:02 fresh kernel: [<c0105792>] setup_frame+0xdd/0x27a
> Dec 29 15:51:02 fresh kernel: [<c0106a58>] common_interrupt+0x18/0x20
> Dec 29 15:51:02 fresh kernel: [<c0105d8e>] handle_signal+0x76/0xc8
> Dec 29 15:51:02 fresh kernel: [<c0105e59>] do_signal+0x79/0xcc
> Dec 29 15:51:02 fresh kernel: [<c013ba62>] mlock_fixup+0x76/0x8c
> Dec 29 15:51:02 fresh kernel: [<c013bacc>] do_mlock+0x54/0x99
> Dec 29 15:51:02 fresh kernel: [<c0105ed3>] do_notify_resume+0x27/0x38
> Dec 29 15:51:02 fresh kernel: [<c01060be>] work_notifysig+0x13/0x15
> Dec 29 15:51:02 fresh kernel: memtest: page allocation failure.
> order:0, mode:0xd2
> Dec 29 15:51:02 fresh kernel: [<c0132b5e>] __alloc_pages+0x29f/0x2bc
> Dec 29 15:51:02 fresh kernel: [<c01405ea>] read_swap_cache_async+0x54/0xba
> Dec 29 15:51:02 fresh kernel: [<c013ae78>] do_swap_page+0x6b/0x22a
> Dec 29 15:51:02 fresh kernel: [<c013b5ae>] handle_mm_fault+0x92/0x125
> Dec 29 15:51:02 fresh kernel: [<c01150fe>] do_page_fault+0x19a/0x4ec
> Dec 29 15:51:02 fresh kernel: [<c01196cc>] call_console_drivers+0xe3/0xeb
> Dec 29 15:51:02 fresh kernel: [<c0119949>] release_console_sem+0x4a/0xc0
> Dec 29 15:51:02 fresh kernel: [<c011989b>] vprintk+0xfa/0x11a
> Dec 29 15:51:02 fresh kernel: [<c011666f>] scheduler_tick+0x317/0x3df
> Dec 29 15:51:02 fresh kernel: [<c0106d68>] print_context_stack+0x18/0x52
> Dec 29 15:51:02 fresh kernel: [<c0114f64>] do_page_fault+0x0/0x4ec
> Dec 29 15:51:02 fresh kernel: [<c0106af5>] error_code+0x2d/0x38
> Dec 29 15:51:02 fresh kernel: [<c019dea0>] __copy_to_user_ll+0x196/0x1b6
> Dec 29 15:51:02 fresh kernel: [<c011d260>] __do_softirq+0x34/0x79
> Dec 29 15:51:02 fresh kernel: [<c0105792>] setup_frame+0xdd/0x27a
> Dec 29 15:51:02 fresh kernel: [<c0106a58>] common_interrupt+0x18/0x20
> Dec 29 15:51:02 fresh kernel: [<c0105d8e>] handle_signal+0x76/0xc8
> Dec 29 15:51:02 fresh kernel: [<c0105e59>] do_signal+0x79/0xcc
> Dec 29 15:51:02 fresh kernel: [<c013ba62>] mlock_fixup+0x76/0x8c
> Dec 29 15:51:02 fresh kernel: [<c013bacc>] do_mlock+0x54/0x99
> Dec 29 15:51:02 fresh kernel: [<c0105ed3>] do_notify_resume+0x27/0x38
> Dec 29 15:51:02 fresh kernel: [<c01060be>] work_notifysig+0x13/0x15
> Dec 29 15:51:02 fresh kernel: VM: killing process memtest
>
>
> I've also ran chkrootkit from unstable to see if it could be anything
> else and to rule other factors out and that can't find any obvious
> breakins.
>
> What I'm curious about why the box will go into an unusable state but
> when restarted it will be fine up until a random point. It could be
> two hours or two days. So if it is the memory why would the box still
> be accepting ICMP ping requests but nothing else that I can see?
>
> Any help in this matter is very much appreciated.
>
> Tam.
The syslog output is the same on my system if I run memtest. This seems to be
some kind of Out-of-Memory condition.
I suggest you should run memtest86 [1] over night to check your memory.
[1] http://www.memtest.org/ or www.knopper.net for a bootable iso-image
http://www.knopper.net/knoppix-mirrors/index-en.html
-Christian
Reply to: