Re: diffrent behaviour of mailq, and some other stuff

[Tomasz Bojakowski <mad@pin.pl>]

Pewnego pięknego dnia 12-Nov-99 Matti Aarnio napisał(-a):
> On Fri, Nov 12, 1999 at 08:43:38AM -0500, Andy Poling wrote:
>> On Fri, 12 Nov 1999, Matti Aarnio wrote:
>> > > As I recall, way back when, the scheduler (and mailq) had an option to
>> > > use a
>> > > UNIX domain socket instead.  Is that no longer an option?
>> [...]
>> >   The code is still there, but it is based on some strange SIGUSR2
>> >   signal handler, and a rendezvous PIPE  (a named pipe)  parametrized
>> >   with a '-q' option to the scheduler.
>> 
>> Ah, yes - that's what I was remembering.
> 
>  Now at CVS (and i even works with AF_UNIX socket!)
> 
>       * proto/scheduler.conf.in, scheduler/mailq.c, scheduler/mq2auth.c,
>         scheduler/prototypes.h, scheduler/readconfig.c,
>         scheduler/scheduler.c, scheduler/transport.c:
>           * scheduler.conf file can now have:
>               PARAMmailqpath = "UNIX:/path/to/af_unix/mailq/port"
>               PARAMmailqpath = "TCP:174"
>           * mailq -p /path/to/af_unix/mailq/port
>           * mailq -p mailq    (check /etc/services for tcp/mailq)
>           * mailq -p 174      (default case)
> 
>           Those are for running multiple schedulers in the system,
>           and allowing each different mailq port.

Well, it may work, however it seem's that mailq is broken either.
I've DISABLED second process of scheduler and roiuter (that which was locking
diffrent postoffice dir). So now zmailer runs like in normal enviroment.

[root@devil|/usr/local/src/zmailer/scheduler]# mailq
Transport queue is empty -- or scheduler uses -Q -mode
[root@devil|/usr/local/src/zmailer/scheduler]# ./mailq -p mailq

Fisrt mailq (which I have in PATH, is taken from 2.99.50s19, the one from
2.99.51 was working in a same way - I mean correct).
It shows the mailq, when mailq is empty it reports it.
 
Now, the second mailq which I've runed is just fetched from cvs, and
compiled... And the previous one (fetched yesterday) was behaving same way
(not reporting anything about mailq...) - when typed mailq, it sadi "nothing",
when I checked /var/spool/postoffice/queue/* there were some messages waiting
(and in a minut they were sent).

I'll be still using that old one mailq, couse it works fine for me (show queue
correctly, even with second scheduler working).


Oh, one more thing: 

[root@devil|/usr/local/src/zmailer/scheduler]# ./mailq         
./mailq: connect failed to 127.0.0.1
./mailq: connect: Connection refused
./mailq: unable to contact scheduler on localhost

[root@devil|/usr/local/src/zmailer/scheduler]# telnet 127.0.0.1 174
Trying 127.0.0.1...
Connected to 127.0.0.1.
Escape character is '^]'.
version zmailer 1.0
Vertices:
End:
Kids: 0  Idle:  0  Msgs:   0  Thrds:   0  Rcpnts:    0  Uptime: 10m11s
Msgs in 0 out 0 stored 0 Rcpnts in 0 out 0 stored 0

hum ? strange... or is new mailq by default accessing UNIX-domain sockets ?
(I'm not too familiar with this subject, still trying to read sth. about
unix-domain sockets)

        Tomasz Bojakowski

/================="Don't get mad, get even."=================\
||              Tomasz "Mad Karrde" Bojakowski              ||
||   tel. +48603081163 , mad@pin.pl , mad@anakata.art.pl    || 
\============================================================/