flexfax: Network problems?

[Alexandre Meissonnier <ameissonnier@dohle.com>]

My configuration: 
- Fax server: 
  Linux Kernel 2.0.32 (S.u.S.E. 5.0.2)
  HylaFAX 4.0pl1
  Modem ZyXEL U1496E/V 6.01 P
- Client machine:
  SCO OpenServer 5.0.2
  HylaFAX 4.0pl1 client software

On the client machine, we run a program once a week that produces about 70 one-page postscript documents (about 1K size) and forks a sendfax command after each document. The whole thing takes only a few minutes .
After the job I verify the log file with the standard output of the sendfax commands: Maybe a dozen lines, not more. I verify the fax server and get the following result:

uucp@helena:/var/spool/fax > ps -x|grep fax
  163  ?  S      0:00 /usr/sbin/faxq 
  165  ?  S      0:00 /usr/lib/fax/hfaxd -i hylafax -o 4557 -s 444 
  172  ?  S      0:00 /usr/lib/fax/faxgetty /dev/ttyS1 
30390  ?  S      0:00 faxsend -m ttyS1 sendq/q746 
uucp@helena:/var/spool/fax > faxstat -s
Can not reach server at host "localhost", port 4559.

The fax server processes are running, but I can't get any connection to the hylafax (=4559)  port. On the client side I can see one sendfax process running. After waiting some minutes, suddenly everything works again. The faxstat command shows that the server is busy sending faxes and new sendfax commands are possible, too. But faxstat -s shows only roughly half the faxes produced by the program in the queue. The rest has vanished.

Questions: 
- Why did some of the sendfax commands simply do NOTHING (not even an error message)?
- Why was the server "dead" for several minutes? Can this be due to too many sendfax commands at the same time? 
I don't believe in a general network problem, since I could login from another machine and even establish an X windows connection (xterm) to the fax server. 

Any ideas?

	Alexandre

--
-------------------------------------------------------------------------------
Alexandre Meissonnier
Dohle Handelsgruppe Systemberatung GmbH
53719 Siegburg (Germany)
ameissonnier@dohle.com