Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

[hylafax-users] t38modem and COMREC invalid response received to PPS

399 views
Skip to first unread message

George H

unread,
Jun 15, 2011, 11:44:33 AM6/15/11
to hylafa...@hylafax.org
Hi list.

I am using hylafax 6.0.4 for debian (hylafax-server_2%3a6.0.4~rc1_i386.deb)
and I have it configured through a t38modem through our company's cisco FOIP
gateway. Receiving faxes works well, but when sending faxes has some
problems. On the cisco end everything seems normal with no errors, on the
hylafax end I keep getting "COMREC invalid response received to PPS" and on
the other side the fax machine complains of missing pages. Sometimes 1 out
of 2 pages gets through and after many re-tries they both arrive.

I am attaching all the logs.
Also what is weird is that when it fails I get an email response saying that
it was completed with non-existent duration time. It should come back as
failed no?

I know as much that COMREC means that the response that hylafax is expecting
it different than what the fax machine is sending... like being out of sync.
But that's as much as I know and I don't know what to do about it or pin
point the problem.

Any help is appreciated.
Thanks

--
George
georg...@gmail.com

email_response.txt
faxstat.txt
hylafax_log.txt
t38config.txt

Lee Howard

unread,
Jun 15, 2011, 4:45:08 PM6/15/11
to George H, hylafa...@hylafax.org
George H wrote:
> I am using hylafax 6.0.4 for debian
> (hylafax-server_2%3a6.0.4~rc1_i386.deb) and I have it configured
> through a t38modem through our company's cisco FOIP gateway. Receiving
> faxes works well, but when sending faxes has some problems. On the
> cisco end everything seems normal with no errors, on the hylafax end I
> keep getting "COMREC invalid response received to PPS" and on the
> other side the fax machine complains of missing pages. Sometimes 1 out
> of 2 pages gets through and after many re-tries they both arrive.
>
> I am attaching all the logs.

We'll need to see the session log to help you much. Those are in
/var/spool/hylafax/log.

> Also what is weird is that when it fails I get an email response
> saying that it was completed with non-existent duration time. It
> should come back as failed no?

It's impossible to tell if the notification message you attached
corresponds to the failure session or not because you did not include
the raw e-mail message or any mail timestamps. Furthermore, it would be
best to see all logs (the notification message, the session log, and the
xferfaxlog entry) corresponding for the call for us to determine whether
the notification message is incorrect.

Thanks,

Lee.


____________________ HylaFAX(tm) Users Mailing List _______________________
To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi
On UNIX: mail -s unsubscribe hylafax-us...@hylafax.org < /dev/null
*To learn about commercial HylaFAX(tm) support, mail sa...@ifax.com.*

George H

unread,
Jun 16, 2011, 1:01:32 AM6/16/11
to Lee Howard, hylafa...@hylafax.org
On Wed, Jun 15, 2011 at 11:44 PM, Lee Howard <fax...@howardsilvan.com>wrote:

> George H wrote:
>
>> I am using hylafax 6.0.4 for debian
>> (hylafax-server_2%3a6.0.4~rc1_i386.deb) and I have it configured through a
>> t38modem through our company's cisco FOIP gateway. Receiving faxes works
>> well, but when sending faxes has some problems. On the cisco end everything
>> seems normal with no errors, on the hylafax end I keep getting "COMREC
>> invalid response received to PPS" and on the other side the fax machine
>> complains of missing pages. Sometimes 1 out of 2 pages gets through and
>> after many re-tries they both arrive.
>>
>> I am attaching all the logs.
>>
>
> We'll need to see the session log to help you much. Those are in
> /var/spool/hylafax/log.
>
>
> Also what is weird is that when it fails I get an email response saying
>> that it was completed with non-existent duration time. It should come back
>> as failed no?
>>
>
> It's impossible to tell if the notification message you attached
> corresponds to the failure session or not because you did not include the
> raw e-mail message or any mail timestamps. Furthermore, it would be best to
> see all logs (the notification message, the session log, and the xferfaxlog
> entry) corresponding for the call for us to determine whether the
> notification message is incorrect.
>
> Thanks,
>
> Lee.
>
>

Hi, thanks. I totally forgot to attach the session log. Sorry about that,
here it is.

--
George
georg...@gmail.com

xferfax.txt

Lee Howard

unread,
Jun 16, 2011, 11:49:42 AM6/16/11
to George H, hylafa...@hylafax.org
George H wrote:
> Jun 15 18:24:45.95: [ 9347]: SEND send PPS (partial page signal)
> Jun 15 18:24:45.95: [ 9347]: SEND send EOP (no more pages or documents)
> Jun 15 18:24:45.95: [ 9347]: <-- [9:AT+FRH=3\r]
> Jun 15 18:24:46.84: [ 9347]: --> [7:CONNECT]
> Jun 15 18:24:47.75: [ 9347]: --> HDLC<5:FF C8 5F C8 E7>
> Jun 15 18:24:47.75: [ 9347]: --> [2:OK]
> Jun 15 18:24:47.75: [ 9347]: SEND recv DCN (disconnect)
> Jun 15 18:24:47.75: [ 9347]: <-- [9:AT+FRS=7\r]
> Jun 15 18:24:47.82: [ 9347]: --> [2:OK]
> Jun 15 18:24:47.82: [ 9347]: COMREC invalid response received to PPS.

As long as your T.38 gateway is not deceiving us (and that is certainly
possible) the session log basically is telling us that the receiver got
the full ECM block of image data and then refused to confirm it and saw
no point in requesting retransmissions.

The only explanation that I can think of would be that the image data is
corrupt and after getting a proper checksummed block the receiver finds
that the image data is corrupt, and so forces a disconnect rather than
confirming the block receipt.

You could test this theory by disabling ModemSoftRTFCC in your modem
config file and then resending with 'sendfax -1' to force MH data to be
sent (instead of MMR). My initial guess would be that there is
something wrong with the MMR encoding. I know that I've fixed some
related things in HylaFAX+ that are not in the hylafax.org releases, and
so another thing to test would be to try HylaFAX+ 5.5.0.

However, after all of this please recognize that it may be the T.38
gateway causing the problems.

George H

unread,
Jun 17, 2011, 6:38:50 AM6/17/11
to Lee Howard, hylafa...@hylafax.org

Lee, always thanks for your replies.

I tried the test you mentioned.
I turned off ModemSoftRTFCC on my T38 modem config.
I sent 'sendfax -1 -n -d 5555 /root/test.txt' and it was sent without
problems. I attached the session log.

With your expert opinion, what is your suggested solution? Do I leave
ModemSoftRTFCC off ? or is there some better solution.

I don't think I will be able to use hylafax+, even though I loved using it
on my Gentoo systems (I was maintaining the ebuild for a few years) but now
we're using Ubuntu distros and I can't seem to find anyone maintaining
hylafax+ .debs.

Thanks and keep up the great work and support!
--
George
georg...@gmail.com

faxlog_new.txt

Lee Howard

unread,
Jun 17, 2011, 4:17:01 PM6/17/11
to George H, hylafa...@hylafax.org
George H wrote:
> On Thu, Jun 16, 2011 at 6:49 PM, Lee Howard <fax...@howardsilvan.com
> <mailto:fax...@howardsilvan.com>> wrote:
>
>
> The only explanation that I can think of would be that the image
> data is corrupt and after getting a proper checksummed block the
> receiver finds that the image data is corrupt, and so forces a
> disconnect rather than confirming the block receipt.
>
> You could test this theory by disabling ModemSoftRTFCC in your
> modem config file and then resending with 'sendfax -1' to force MH
> data to be sent (instead of MMR). My initial guess would be that
> there is something wrong with the MMR encoding. I know that I've
> fixed some related things in HylaFAX+ that are not in the
> hylafax.org <http://hylafax.org> releases, and so another thing to

> test would be to try HylaFAX+ 5.5.0.
>
>
>
> I tried the test you mentioned.
> I turned off ModemSoftRTFCC on my T38 modem config.
> I sent 'sendfax -1 -n -d 5555 /root/test.txt' and it was sent without
> problems. I attached the session log.
>
> With your expert opinion, what is your suggested solution? Do I leave
> ModemSoftRTFCC off ? or is there some better solution.

So it would seem that either the T.38 gateway is fouling-up the MMR
image data (which seems somewhat unlikely) or the version of HylaFAX
you're using has some problems corrupting MMR image data (which is what
I suspect is the case). The solution to the latter would be to use
HylaFAX+ 5.5.0. If the problem is the former, then you should probably
just avoid using MMR (which would be to leave ModemSoftRTFCC off and
always submit faxes with 'sendfax -1' or equivalent).

> I don't think I will be able to use hylafax+, even though I loved
> using it on my Gentoo systems (I was maintaining the ebuild for a few
> years) but now we're using Ubuntu distros and I can't seem to find
> anyone maintaining hylafax+ .debs.

I maintain some HylaFAX+ deb files, and they're on the Sourceforge site
download pages. However, it's easy-enough to build your own deb file
(and this is what I'd recommend if you don't want to do a source-code
installation) using the "hylafax debian build files" which are also
downloadable from Sourceforge. (Read the README file within that tarball.)

George H

unread,
Jun 17, 2011, 4:26:29 PM6/17/11
to Lee Howard, hylafa...@hylafax.org

Thanks Lee, Just a quick question... why isn't hylafax dropped for the more
superior hylafax+ version ? I also noticed that FaxAccounting script that
worked with hylafax+ doesn't anymore with hylafax.

--
George
georg...@gmail.com

George H

unread,
Jun 20, 2011, 2:55:14 AM6/20/11
to Lee Howard, hylafa...@hylafax.org

> Good question for the Debian maintainer to answer.
>
> Thanks,
>
> Lee.
>

Hi,

I am trying to force hylafax to send faxes with the -1 parameter until I
have time to upgrade all the servers to hylafax+, but I am having some
difficulties doing so.

I added "FaxSendCmd: /usr/sbin/sendfax -1" line to
/var/spool/hylafax/config and it doesn't seem to take the parameter. I also
read the man page of faxconfig program but it doesn't seem to configure any
of the parameters I want.

Is there a way to do this?

Thanks.
--
George
georg...@gmail.com

Lee Howard

unread,
Jun 21, 2011, 1:21:52 AM6/21/11
to George H, hylafa...@hylafax.org
George H wrote:
> I am trying to force hylafax to send faxes with the -1 parameter until
> I have time to upgrade all the servers to hylafax+, but I am having
> some difficulties doing so.
>
> I added "FaxSendCmd: /usr/sbin/sendfax -1" line to
> /var/spool/hylafax/config and it doesn't seem to take the parameter. I
> also read the man page of faxconfig program but it doesn't seem to
> configure any of the parameters I want.
>
> Is there a way to do this?

FaxSendCmd refers to faxsend and not sendfax.

How do you submit faxes? Normally you'd tell your client program to use MH.

But you can probably do it with "Use2D: no" in /var/spool/hylafax/config.

George H

unread,
Jun 21, 2011, 1:42:32 AM6/21/11
to Lee Howard, hylafa...@hylafax.org
On Tue, Jun 21, 2011 at 8:21 AM, Lee Howard <fax...@howardsilvan.com> wrote:

> George H wrote:
>
>> I am trying to force hylafax to send faxes with the -1 parameter until I
>> have time to upgrade all the servers to hylafax+, but I am having some
>> difficulties doing so.
>>
>> I added "FaxSendCmd: /usr/sbin/sendfax -1" line to
>> /var/spool/hylafax/config and it doesn't seem to take the parameter. I also
>> read the man page of faxconfig program but it doesn't seem to configure any
>> of the parameters I want.
>>
>> Is there a way to do this?
>>
>
> FaxSendCmd refers to faxsend and not sendfax.
>
> How do you submit faxes? Normally you'd tell your client program to use
> MH.
>
> But you can probably do it with "Use2D: no" in /var/spool/hylafax/config.
>
> Thanks,
>
> Lee.
>


Ah I didn't realize the faxsend / sendfax.
I was also scanning the man pages for MMR and MH but didn't find anything.
The Use2D param did the trink.

Thanks Lee.
Now I'll try to package hylafax+ as a .deb and upgrade these servers.

--
George
georg...@gmail.com

0 new messages