On Thu, Jun 22, 2006 at 05:12:22PM -0300, Marc G. Fournier
wrote:>
> 'k, I was thinking / hoping it was just maxpipekva causing me issues,
but
> apparently that isn't it ... it just froze up again with only 40 out of
> 60M of allocated space used:
>
> =============================> kern.ipc.maxpipekva: 67108864 -
kern.ipc.pipekva: 39272448
> running processes: 1631
>
> =============================>
> ps shows processes are still running:
>
> last pid: 84729; load averages: 0.29, 0.61, 3.972 up 0+04:42:36
> 17:07:12
> 3657 processes:3656 sleeping, 1 zombie
> CPU states: 1.3% user, 0.0% nice, 2.9% system, 0.2% interrupt, 95.7%
> idle
> Mem: 3105M Active, 341M Inact, 350M Wired, 118M Cache, 112M Buf, 24M Free
> Swap: 8192M Total, 728M Used, 7464M Free, 8% Inuse, 24K In
> Number of processes to show:
> PID USERNAME THR PRI NICE SIZE RES STATE C TIME WCPU COMMAND
> 34702 www 1 20 0 22552K 14832K lockf 0 0:14 0.78% httpd
> 15113 www 1 4 0 23980K 12944K sbwait 0 0:02 0.44% httpd
> 8903 www 1 20 0 22312K 14560K lockf 0 0:14 0.34% httpd
> 25356 www 1 4 0 18236K 7352K sbwait 1 0:10 0.10% httpd
> 6324 60 1 4 0 3592K 1068K select 0 9:45 0.00% master
> 6204 root 1 4 0 1300K 680K select 1 4:52 0.00% syslogd
> 10824 www 1 96 0 754M 144M select 0 4:03 0.00% java
> 22816 www 13 20 0 636M 65200K kserel 1 3:08 0.00% java
> 24527 501 15 20 0 202M 159M kserel 0 1:45 0.00% nsd
> 25038 nobody 13 20 0 151M 117M kserel 1 1:12 0.00% nsd
> 28812 ipaudit 14 20 0 160M 126M kserel 1 1:10 0.00% nsd
> 5191 www 6 20 0 1157M 25428K kserel 0 1:00 0.00% java
> 2945 nobody 13 20 0 131M 84284K kserel 0 0:58 0.00% nsd
> 16014 root 6 20 0 1130M 36852K kserel 0 0:58 0.00% java
> 8938 www 6 20 0 1141M 40336K kserel 0 0:54 0.00% java
> 22535 www 6 20 0 1139M 29412K kserel 0 0:53 0.00% java
> 19295 ipaudit 16 20 0 188M 123M kserel 1 0:51 0.00% nsd
> 19251 88 8 20 0 59340K 5572K kserel 0 0:44 0.00% mysqld
> 17287 ipaudit 14 20 0 83388K 31688K kserel 0 0:43 0.00% nsd
> 17279 ipaudit 13 20 0 107M 67680K kserel 1 0:37 0.00% nsd
> 17243 ipaudit 12 20 0 90740K 52000K kserel 1 0:35 0.00% nsd
> 25487 nobody 22 20 0 85652K 63688K kserel 1 0:33 0.00% nsd
> 31657 nobody 12 20 0 40612K 26380K kserel 1 0:31 0.00% nsd
> 17255 ipaudit 14 20 0 61108K 28044K kserel 0 0:30 0.00% nsd
> 17262 ipaudit 13 20 0 53144K 20444K kserel 1 0:30 0.00% nsd
> 17270 ipaudit 13 20 0 54772K 20524K kserel 0 0:28 0.00% nsd
> 8743 nobody 22 20 0 40984K 11016K kserel 0 0:27 0.00% nsd
> 15110 88 6 20 0 57772K 4720K kserel 1 0:20 0.00% mysqld
> 5651 88 6 20 0 57132K 1652K kserel 1 0:18 0.00% mysqld
> 6323 88 6 20 0 57016K 1652K kserel 1 0:18 0.00% mysqld
>
> I'm at a loss as to what to look at though ... other then the fact that
> I've hit the limit as far as 'overloading the system' is
concerned :(
What exactly happens ? Are you sure that the _system_ freezes ? Could
it be that only your application has some troubles ?
Basically, you did not say whether you could log in the failed machine,
are existing shells responsible, etc.
For reporting kernel deadlocks, see
http://www.freebsd.org/doc/en_US.ISO8859-1/books/developers-handbook/kerneldebug-deadlocks.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url :
http://lists.freebsd.org/pipermail/freebsd-stable/attachments/20060623/44ce05d4/attachment.pgp