Re: Hylafax+-5.5.9 & blocked queues of the same outbound fax number

On 07/21/2018 06:57 PM, Gregory P. Ennis wrote:
> After 'so many' faxes to the same number hylafax+ stops sending to
> that
> number.=C2=A0=C2=A0Other outbound numbers work without a problem, but t=
he fax
> queue with the blocked number remains full of the blocked
> number.=C2=A0=C2=A0When
> this happens I have to restart hylafax and the fax queues of the
> blocked fax start again.
>=20
> It appeared to me that hylafax was recognizing a large queue of the
> same number and interpreted the presence of this large queue as a bad
> fax.=C2=A0=C2=A0The receiving fax is a busy fax and does have some 'bus=
y'
> problems.

So, to investigate this problem it will be necessary to know the=C2=A0
configuration, because various configuration items can have
considerable=C2=A0
effect on what is going on in the queues - especially when dealing
with=C2=A0
fax jobs to the same number.

Are you batching or not?=C2=A0 (What is your MaxBatchJobs setting in your=
=C2=A0
/var/spool/hylafax/etc/config file?)

What is your MaxConcurrentCalls setting?=C2=A0 (It is also in etc/config.=
)

What is your ShareCallFailures setting?=C2=A0 (Likewise in etc/config.)

Are you using ModemGroup at all, and if so, what is the ModemGroup=C2=A0
configuration, and what "modem" is being used in the job submission?

(So, maybe it's best if you just share your etc/config file.)

> I have changed the value of MaxConsecutiveBadCmds:=C2=A0=C2=A060 in
> /etc/hylafax/hfaxd.conf which was originally set to 10.=C2=A0=C2=A0This
> appeared
> to help, but did not solve the problem.

Settings in hfaxd.conf affect the client-server communication and do
not=C2=A0
usually have a role in the faxq operation or fax protocol operation.

In this case, MaxConsecutiveBadCmds influences how tolerant hfaxd is=C2=A0
with your client program.=C2=A0 Did you write your own client program or
are=C2=A0
you using sendfax or faxmail as most users (in which case=C2=A0
MaxConsecutiveBadCmds will not likely do anything)?

> The number of consecutive
> faxes to the same number has been in the range of 40 to 45.
>=20
> The /var/messages entries appear to be :
>=20
> Jul 18 22:58:52 MaBa FaxQueuer[18820]: NOTIFY: bin/notify-smile
> 'sendq/q93617' 'blocked' ''
> Jul 18 22:58:52 MaBa FaxQueuer[18701]: NOTIFY: bin/notify-smile
> 'sendq/q93612' 'requeued' '' '23:03'
> Jul 18 22:58:52 MaBa FaxQueuer[18701]: NOTIFY: bin/notify-smile
> 'sendq/q93613' 'requeued' '' '23:03'
> Jul 18 22:58:52 MaBa FaxQueuer[18701]: NOTIFY: bin/notify-smile
> 'sendq/q93614' 'requeued' ''

This is normal.=C2=A0 What should happen, then, after some time is that i=
t=C2=A0
should reattempt those jobs.=C2=A0 It sounds like you're seeing it not do=
=C2=A0
that, but rather they stay "blocked" forever or they simply stay stuck=C2=
=A0
and are never reattempted.

Ultimately I may need you to set ServerTracing (in etc/config) to
0xFFF,=C2=A0
restart faxq, and then let it run again until a problem happens and
then=C2=A0
provide me the FaxQueuer entries from your syslog.

And, it would be helpful to know if 'faxalter -p <jobid>' works to get=C2=
=A0
the "stuck" jobs to go out.

Thanks,

Lee.

---------------------------------------------------------------

Lee,

Thank you for helping me debug this problem !!!!!
To answer your questions.

We are not batching, although I thought perhaps we should consider
doing that.  The problem that I see is that these 40 to 50 commands
each have a different image to the same number.  We are using a cover
page so each command has two pages. =20

Here is the content of my /var/spool/hylafax/etc/config file :

LogFacility:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0daemon
CountryCode:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A01
AreaCode:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=
=C2=A0=C2=A0=C2=A0=C2=A0972
LongDistancePrefix:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A01
InternationalPrefix:=C2=A0=C2=A0=C2=A0=C2=A0011
DialStringRules:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0etc/dialr=
ules
ServerTracing:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=
1
ModemGroup:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0y
NotifyCmd:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0bin/notify-smile

As you can see I have not set any of the parameters that you suggested.
   but "ModemGroup" was set to "1" by default and I left that in place,
but I have not set a preference for the modems, and my perception is
that the 'round robin' method is used by hylfax.  When I watch the
queues that go out to the same number hylfax alternates the outbound
modems after each transmission. =20

I am using sendfax for all submissions; in the case of the jobs that
are getting stuck, I am using sendfax on a machine inside the network,
but different than the machine with the fax modems.

I had already set ServerTracing to 0xFFF in /etc/hylafax/hfad.config

After reviewing your note and doing some reading I have modified the
config file so that it is now :

LogFacility:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0daemon
CountryCode:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A01
AreaCode:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=
=C2=A0=C2=A0=C2=A0=C2=A0972
LongDistancePrefix:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A01
InternationalPrefix:=C2=A0=C2=A0=C2=A0=C2=A0011
DialStringRules:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0etc/dialr=
ules
ServerTracing:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=
1
ModemGroup:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0y
NotifyCmd:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0bin/notify-smile
SessionTracing:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A00xFF=
F
MaxConcurrentCalls:=C2=A0=C2=A0=C2=A0=C2=A0=C2=A060


We are scheduled to have about 40 to 50 of these faxes go out tomorrow
I will use 'faxalter -p <jobid>' and will let you know what I find.

Thanks again for your help, and thanks so much for your work on
hylafax; this is the first real problem I have had with it in the last
5 years.

Greg
'


____________________ HylaFAX(tm) Users Mailing List _____________________=
__
  To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.c=
gi
 On UNIX: mail -s unsubscribe [email protected] < /dev/nu=
ll
  *To learn about commercial HylaFAX(tm) support, mail [email protected].*