[hylafax-users] JKILL and Trigger

["Paolo Bagnoli" <p.bagnoli@xxxxxxxxxxxxxxxxxxxxxx>]

May 18, 2005                   11:19


Hello !
We are using HylaFAX 4.2.1 and Fedora Core 3.
It's very good but have we found a bug?

If we try to kill a job sended to a busy or not existing
fax number while the phone connection is down but the job
is in queue waiting to be resended, the trigger connection
of our application appears frozen and after some time similar
errors "FaxQueuer[4324]: HylaClient::send: client/4374:
write failed (return 560, seq# 4294967295, Resource
temporarily unavailable)" appear in /var/log/messages.

This behaviour is easily reproducible with HylaFAX utilities also:=20
- Run the faxwatch utility with 'faxwatch S*J*R*'.
- Sending a fax with sendfax to a not existing phone number
  (e.g. 0000001122).
  At this moment you can see faxwatch regular messages.
- Wait for the first try to sending job be completed and job become slee=
ping.
  At this moment faxwatch received all regular messages and stand by.
- Then send a JKILL command for the corresponding jobid
  (e.g. with a telnet connection open on port 4559):
  faxwatch trigger seems frozen and no others messagges will be
  displayed even sending or receiving new fax.
We have noted problems in /var/spool/hylafax/client/xxxx FIFO files.
Is it a HylaFAX bug?
Any work around?
P.S. Please, ignore the previous similar post.
Thanks
Paolo Bagnoli



Paolo Bagnoli
p.bagnoli@xxxxxxxxxxxxxxxxxxxxxx