Re: [hylafax-users] Stop and wait failure/No response to PPS

[Lee Howard <faxguy@xxxxxxxxxxxxxxxx>]

Bruce McAlister wrote:

> HylaFAX -> IAXModem -> Asterisk -> SS7 Switch
>
> The above is for sending faxes, for recieving, the reverse applies.
>
> HylaFAX, IAXModem and Asterisk all live on the same host, and the SS7 is
> on a local, low-latency network, the ping responses is sub 1
> millisecond. Output of a sample ping, with IP's removed follows:
>
> --------------------------------------------------------------------------------------------------------
> 64 bytes from SS7 (0.0.0.0): icmp_seq=0. time=0.766 ms
> 64 bytes from SS7 (0.0.0.0): icmp_seq=1. time=0.743 ms
> 64 bytes from SS7 (0.0.0.0): icmp_seq=2. time=0.955 ms

There are numerous people that would love to push fax audio over a 
"local, low-latency network"... some get away with it, some don't.  You 
seem to be doing okay with it... for the most part.  But, beware... ;-)

> Oct 02 15:03:10.40: [18573]: <-- [16:ATDT35300000000\r]
> Oct 02 15:03:20.83: [18573]: --> [7:CONNECT]
> Oct 02 15:03:21.21: [18573]: --> [4:9E 00 10 03]
> Oct 02 15:03:21.21: [18573]: --> [5:ERROR]
>  

This is an indication that the audio was corrupt.  It may not be your 
fault, but if you see enough of this then the likelihood that the 
culprit is on your end is greater.

> Oct 02 15:03:50.71: [18573]: <-- [9:AT+FTH=3\r]
> Oct 02 15:03:50.72: [18573]: --> [7:CONNECT]
> Oct 02 15:03:50.72: [18573]: <-- HDLC<7:FF C8 FD F2 00 00 5A>
> Oct 02 15:03:59.36: [18573]: --> [2:OK]
> Oct 02 15:03:59.36: [18573]: SEND send PPS (partial page signal)
> Oct 02 15:03:59.36: [18573]: SEND send MPS (more pages, same document)
> Oct 02 15:03:59.36: [18573]: <-- [9:AT+FRH=3\r]
> Oct 02 15:04:02.47: [18573]: --> [13:AT+FRH=3ERROR]
> Oct 02 15:04:02.47: [18573]: MODEM TIMEOUT: waiting for v.21 carrier
> Oct 02 15:04:02.68: [18573]: MODEM TIMEOUT: reading line from modem
> Oct 02 15:04:02.68: [18573]: MODEM <Timeout>
> Oct 02 15:04:02.68: [18573]: <-- [9:AT+FTH=3\r]
> Oct 02 15:04:02.68: [18573]: --> [8:AT+FTH=3]
> Oct 02 15:04:32.69: [18573]: MODEM TIMEOUT: reading line from modem
>  

Well, the iaxmodem "reset" itself... somehow.  Probably had a 
disconnnection at an inopportune time.  Some of the modem responses 
there seem a bit erronious... but as I see in later logs, I'm betting on 
some kind of iaxmodem usage or configuration error.

> Oct 02 14:10:56.50: [13443]: SEND send PPS (partial page signal)
> Oct 02 14:10:56.50: [13443]: SEND send MPS (more pages, same document)
> Oct 02 14:10:56.50: [13443]: <-- [9:AT+FRH=3\r]
> Oct 02 14:10:56.90: [13443]: --> [7:CONNECT]
> Oct 02 14:10:57.94: [13443]: --> HDLC<5:FF C8 31 45 8F>
> Oct 02 14:10:57.97: [13443]: --> [2:OK]
> Oct 02 14:10:57.97: [13443]: SEND recv MCF (message confirmation)
>  

This is a good sign that the audio... at least during this call... was 
not corrupted.

> Oct 02 14:11:01.26: [13443]: SEND send PPS (partial page signal)
> Oct 02 14:11:01.26: [13443]: SEND send EOP (no more pages or documents)
> Oct 02 14:11:01.26: [13443]: <-- [9:AT+FRH=3\r]
> Oct 02 14:11:01.65: [13443]: --> [7:CONNECT]
> Oct 02 14:11:02.69: [13443]: --> HDLC<5:FF C8 31 45 8F>
> Oct 02 14:11:02.70: [13443]: --> [2:OK]
> Oct 02 14:11:02.70: [13443]: SEND recv MCF (message confirmation)
>  

Again, another good sign.

> Oct 02 14:20:25.63: [13992]: <-- [11:AT+FTM=146\r]
> Oct 02 14:20:25.64: [13992]: --> [7:CONNECT]
> Oct 02 14:20:25.64: [13992]: DELAY 400 ms
> Oct 02 14:21:09.80: [13992]: --> [9:AT+VCID=1]
> Oct 02 14:21:09.80: [13992]: --> [6:ATS0=0]
> Oct 02 14:21:09.80: [13992]: --> [2:OK]
> Oct 02 14:21:09.80: [13992]: MODEM set XON/XOFF/DRAIN: input ignored, output disabled
> Oct 02 14:21:09.80: [13992]: <-- [9:AT+FTS=7\r]
> Oct 02 14:21:09.80: [13992]: --> [8:AT+FTS=7]
> Oct 02 14:21:09.80: [13992]: --> [5:ERROR]

This here is simply bizarre.  I cannot think of a normal circumstance 
that would cause iaxmodem to report "AT+VCID=1" and then "ATS0=0" during 
a Phase C transmit session.  I can, however, believe that it is an 
indication that *HylaFAX* is attempting to reset the modem.

Here's what I think is happening...

You're running faxgetty.  That works fine receiving faxes.  However, 
when faxsend goes about its business faxgetty is supposed to take a 
break and watch faxsend's lockfile on the modem... and when the lockfile 
disappears then it will reinitialize/reset the modem.

So... now it's time to take a look at the modem file locking.  On Linux 
you would generally see "LCK..devicename" lock files appear in a 
directory like /var/lock/.  Do you see that happening?  Do you see it 
happening agreeably between faxgetty and faxsend?

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-users-request@xxxxxxxxxxx < /dev/null
 *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*