jesus at omniti
Oct 19, 2003, 5:03 PM
Post #4 of 6
(998 views)
Permalink
|
My guess is that this is an uninitialized variable.
I will try to run wackamole under valgrind on Linux to track this down
(unless someone beats me to it -- any takers).
On Wednesday, Oct 15, 2003, at 12:42 US/Eastern, Cauthorn, Matt
(MAN-Corporate) wrote:
> Wednesday, October 15, 2003, 12:19:53 PM, Cauthorn, Matt
> (MAN-Corporate)
> wrote:
>
> CMMC> We're having some problems with Wackamole on FreeBSD 5.1...the
> setup
> is very
> CMMC> simple (nearly default, in fact):
>
> CMMC> Basic spread config:
> CMMC> Spread_Segment 10.0.0.255:4803 {
>
> CMMC> qmail1b 10.0.0.1
> CMMC> qmail2b 10.0.0.2
> CMMC> }
>
> CMMC> -- Spread is up and running. Works great.
>
> CMMC> Wackamole.conf:
>
> CMMC> Spread = 4803
> CMMC> SpreadRetryInterval = 5s
> CMMC> Group = wack1
> CMMC> # Named socket for online control
> CMMC> Control = /var/run/wack.it
> CMMC> Prefer None
> CMMC> VirtualInterfaces {
>
> CMMC> { bge1:10.0.0.10/32 }
> CMMC> }
> CMMC> Arp-Cache = 90s
>
> CMMC> Notify {
>
> CMMC> #bge1:10.0.0.1/32
> CMMC> arp-cache
> CMMC> }
> CMMC> balance {
>
> CMMC> AcquisitionsPerRound = all
> CMMC> interval = 4s
> CMMC> }
> CMMC> # How long it takes us to mature
> CMMC> mature = 5s
>
> CMMC> The symptoms are:
>
> CMMC> 1) Start spread. Comes right up.
> CMMC> 2) Start Wackamole...get the boiler plate message, with the
> credits,
> etc.,
> CMMC> but no daemon. We've got pid files and the socket setup in
> /var/run/,
> CMMC> however the process id doesn't exist at all. It doesn't give us
> any
> errors
> CMMC> either. Silent.
>
> CMMC> Unfortunately, ktrace doesn't seem to give us too much more than
> this.
> The
> CMMC> last few lines of a ktrace.out shows this:
>
> CMMC> 60751 wackamole RET write 81/0x51
> CMMC> 60751 wackamole CALL fork
> CMMC> 60751 wackamole RET fork 60752/0xed50
> CMMC> 60751 wackamole CALL exit(0x1)
>
> CMMC> ....which may or may not be meaningful - I just don't know. No
> other
> obvious
> CMMC> errors were noticed in the trace.
>
> CMMC> We mounted /proc, as a hunch, but that doesn't seem to work
> either.
> Our
> CMMC> FreeBSD 4.8 machine works well wackamole-wise. Any ideas would be
> CMMC> appreciated.
>
> CMMC> Regards,
> CMMC> Matt
>
>
> CMMC> _______________________________________________
> CMMC> wackamole-users mailing list
> CMMC> wackamole-users[at]lists.backhand.org
> CMMC> http://lists.backhand.org/mailman/listinfo/wackamole-users
>
>
> Hi Matt,
> If you want to see wackamole printing to the screen, run it with -d
> i.e. in debug mode. Wackamole is supposed to run as a daemon, so it
> forks of a daemon process and quits.
>
>
>
>
> Ashima mailto:munjal[at]cnds.jhu.edu
> -----------------------------------
>
>
> Ok - Thanks for the tip. Now we're seeing this from ktrace:
>
> ##########################
>
> 60798 wackamole GIO fd 2 wrote 31 bytes
> "socket: Protocol not supported
> "
> 60798 wackamole RET writev 31/0x1f
> 60798 wackamole CALL exit(0x1)
>
> This is dumped to the screen as well ( Protocol not supported ).
>
> Matt
>
> _______________________________________________
> wackamole-users mailing list
> wackamole-users[at]lists.backhand.org
> http://lists.backhand.org/mailman/listinfo/wackamole-users
>
> _______________________________________________
> wackamole-users mailing list
> wackamole-users[at]lists.backhand.org
> http://lists.backhand.org/mailman/listinfo/wackamole-users
>
// Theo Schlossnagle
// Principal Engineer -- http://www.omniti.com/~jesus/
// Postal Engine -- http://www.postalengine.com/
// Ecelerity: fastest MTA on earth
|
