| 1 |
Brian Kroth wrote: |
| 2 |
> Brian Kroth wrote: |
| 3 |
>> Natanael Copa wrote: |
| 4 |
>>> On Sun, 2007-11-04 at 01:34 +0200, pageexec@××××××××.hu wrote: |
| 5 |
>>>> On 3 Nov 2007 at 19:02, Brian Kroth wrote: |
| 6 |
>>> |
| 7 |
>>>>> PS - would you like me to take this off list? |
| 8 |
>>>> no need unless you don't want to leak any more of your host names |
| 9 |
>>>> in kernel logs ;-). actually, attachments this big you can send me |
| 10 |
>>>> directly and don't bother the list with it, but the discussion may |
| 11 |
>>>> be interesting for others as well. |
| 12 |
>>>> |
| 13 |
>>> |
| 14 |
>>> This discussion is interesting, yes |
| 15 |
>>> |
| 16 |
>>> -nc |
| 17 |
>>> |
| 18 |
>> |
| 19 |
>> Getting more interesting. It just rebooted itself, which it hasn't |
| 20 |
>> done yet. Here's the two messages in kern.log before and after it |
| 21 |
>> happened. Nothing unexpected in any of the other log files either. |
| 22 |
> |
| 23 |
> Actually, it looks like ipmi watchdog caused it. |
| 24 |
> |
| 25 |
> # ipmitool sel list |
| 26 |
> |
| 27 |
> 1a | 11/05/2007 | 02:24:22 | Watchdog 2 #0x50 | Hard reset | Asserted |
| 28 |
> 1b | 11/05/2007 | 11:12:39 | Watchdog 2 #0x50 | Hard reset | Asserted |
| 29 |
> |
| 30 |
> Still not quite sure why that would have happened yet. |
| 31 |
> |
| 32 |
> pagexec: were those pax messages any help to you? |
| 33 |
> |
| 34 |
> Thanks, |
| 35 |
> Brian |
| 36 |
|
| 37 |
Another update. I disabled IPMI in order to see what was causing that. |
| 38 |
The machine with 2.6.23-hardened-r1_patched panicked and locked up. I |
| 39 |
couldn't get the message though. |
| 40 |
|
| 41 |
I need it up and running so I moved it back to 2.6.20-hardened-10, which |
| 42 |
works for me, and am working on setting up a VMWare image of the machine |
| 43 |
in order to reproduce the problem and test it there. |
| 44 |
|
| 45 |
Also, I had another machine complain about a bad page state. This one |
| 46 |
was still using the original kernel config that I stared having the bad |
| 47 |
page state messages so there's not many details there. However, the |
| 48 |
program that triggered it is from a ColdFusion binary install and gives |
| 49 |
the following from paxctl: |
| 50 |
|
| 51 |
technetium ~ # paxctl -v /opt/coldfusionmx7/verity/k2/_ilnx21/bin/k2admin |
| 52 |
PaX control v0.4 |
| 53 |
Copyright 2004,2005,2006 PaX Team <pageexec@××××××××.hu> |
| 54 |
|
| 55 |
file /opt/coldfusionmx7/verity/k2/_ilnx21/bin/k2admin does not have a |
| 56 |
PT_PAX_FLAGS program header, try conversion |
| 57 |
|
| 58 |
Nov 6 11:18:53 technetium Bad page state in process 'k2admin' |
| 59 |
Nov 6 11:18:53 technetium page:c1342560 flags:0x40000001 |
| 60 |
mapping:00000000 mapcount:0 count:0 |
| 61 |
Nov 6 11:18:53 technetium Trying to fix it up, but a reboot is needed |
| 62 |
Nov 6 11:18:53 technetium Backtrace: |
| 63 |
Nov 6 11:18:53 technetium [<c044a16f>] [<c044ab14>] [<c044b26c>] |
| 64 |
[<c044d928>] [<c045c18a>] [<c0452827>] [<c045296d>] [<c0452ac6>] |
| 65 |
[<c045826c>] [<c041aabc>] [<c0469bea>] [<c046a008>] [<c046a387>] |
| 66 |
[<c0469ab2>] [<c0494ce2>] [<c044aeb9>] [<c053ba06>] [<c0469579>] |
| 67 |
[<c046a880>] [<c046ac39>] [<c0401617>] [<c0402a72>] |
| 68 |
======================= |
| 69 |
|
| 70 |
Brian |
Archives