On 2007-Mar-06 14:46:50 +1030, Daniel O'Connor <[EMAIL PROTECTED]> wrote:
>Would a crash dump be useful? I think I will be able to update the ataraid
>stuff to allow a dump onto the array.
I doubt it will be feasible to make any progress without either a
crashdump or KDB/DDB access to the system.
On Tuesday 06 March 2007 05:04, Peter Jeremy wrote:
> How difficult would it be to build a test system somewhere where the
> console was accessible? I don't think you are going to make progress
> without console access.
Not possible, I can't move the system as it is very remote and there are no
On 2007-Mar-05 11:36:41 +1030, Daniel O'Connor <[EMAIL PROTECTED]> wrote:
>I was incrementally dumping sysctl trees and when I got to vm it hung..
>
>eureka:~>sysctl vm
>load: 0.07 cmd: sysctl 72864 [user map] 0.00u 0.00s 0% 852k
>load: 0.07 cmd: sysctl 72864 [user map] 0.00u 0.00s 0% 852k
>load:
On Monday 05 March 2007 11:36, Daniel O'Connor wrote:
> I tried to reboot in another window but..
> eureka:~>reboot
> load: 0.06 cmd: csh 71159 [allproc] 0.03u 0.00s 0% 4652k
Interestingly this process did eventually finish. The system hasn't rebooted
yet but perhaps it will eventually.
--
Dan
On Thursday 01 March 2007 17:35, Daniel O'Connor wrote:
> > in /rescue and use './sysctl' (and other commands in rescue). I think
> > you will still be able to execute static executables in the current
> > directory vis a relative path even if the FS is deadlocked. (As long
> > as your shell isn'
On Thursday 01 March 2007 17:14, Peter Jeremy wrote:
> >The problem seems to occur when the acquisition process (called the
> > Recorder) tries to fork off a new child.
>
> Your description and the ps output suggest a filesystem deadlock.
When it happened last time I had no problem running command
On 2007-Mar-01 11:48:54 +1030, Daniel O'Connor <[EMAIL PROTECTED]> wrote:
>The problem seems to occur when the acquisition process (called the Recorder)
>tries to fork off a new child.
Your description and the ps output suggest a filesystem deadlock.
>Does anyone have any suggestions for debuggin
We run a set of meteor and wind radar systems based on FreeBSD and one of them
I am seeing an odd problem that does not occur very often (once a month
perhaps).
The main process we use for gathering data regularly fork()s children off to
process data (at each 30 minute acquisition cycle). The data
