[cisco-voip] Question regarding Rightfax Configuration and "Transmission Error"

[Frank Arrasmith]

On average or failure rate is below 5%, however there are some senders with
failure rates as high as 80%. We usually reach out to them and ask them to
turn off ECM and some transmit speed to 9600.  In response to some of the
comments.

We have MGCP and SIP internally with 99.9% success rate.

We have both SIP to PRI  and SIP to SIP for external. Carrier is Level 3.
We used to have AT&T but  had the same problem.

Outbound faxing from the Rightfax server is fine. It's only with inbound
faxing that we have this problem. This is why I'm questioning a server
setting. It doesn't make sense to me that the server would toss out the fax
just because there was a single RTN message sent, especially when the
packet capture shows the sending device complied with the message, slowed
it's transmit speed and finished the fax without any further issue.

On the positive side, working on this over the past year has really helped
me to build my FoIP skills, and faxing within the test of our network is
rock solid 😀


On Thu, Apr 6, 2017 at 7:04 AM Scott Voll <svoll.voip at gmail.com> wrote:

> Wow Frank, that is a lot of troubleshooting, and a great run down.
>
> Can I ask what your failure rate is?
>
> We have about a 2-4% failure rate and have decided not to troubleshoot
> those as they tend to be a Telco issue between us and the sender.  We use a
> lot of toll bypass so if it doesn't work over a pri then we send it out
> another one.  Usually fixes most of our problems ;-)
>
> Scott
>
>
> On Wed, Apr 5, 2017 at 5:06 PM, Frank Arrasmith <frank.arrasmith at gmail.com
> > wrote:
>
> Calling all Rightfax gurus,
>    I have the following question regarding the Rightfax configuration and
> transmission errors.
>
> Background:
> My enterprise has CUCM 10.5 with a pretty dialed in Fax/T38 setup over SIP
> and MGCP gateways.  For the most part, faxing is pretty solid inbound and
> outbound to and from PSTN GW's, CUBEs, ,analog VG's, and regular fax
> machines(we have a lot).  We have a SIP trunk connected to Rightfax 10.5
> server, which is managed by another group where we have limited
> access/experience with Rightfax configuration settings.  It took us awhile
> to get the Rightfax servers with the correct t38 setup because they had
> only run traditional CAS T1 prior to us, so it was new to everyone, but it
> is up and stable except for the following issue..
>
> Symptom :
> The problem we see with Rightfax is with Transmission errors.  It starts
> with our internal customers reporting that they do not receive a fax even
> when the sender receives confirmation.  Upon further review we see that the
> suspect call is listed as a "Transmission Error" in Rightfax, so the fax
> never gets delivered to the customers account/mailbox even though the call
> completes.
>
> Analysis:
> Since we are running T38, we can packet capture at the server, and  we see
> normal fax protocol exchange, except for the suspect calls where we see
> "RTN" Messages.  My understanding of the T.30 protocol is that when a RTN
> message is delivered to the sender, that is an indication for the sender to
> slow down and resend the last page. We actually see in the messages where
> the RTN message gets sent, and the sender complies with the notice, and
> sends again at 12000, then call completes as normal with an EOP message and
> a DCN message.  In these cases,the Rightfax team actually looks at the fax
> image and may see a bit of blurriness, but perfectly legible text ,even tho
> its still marked as transmission failure. We have asked them if there is a
> setting that can be tuned where the RTN does not cause the service to mark
> the transmission as failure. To this they reply, "It was working fine
> before when we were on CAS, so it must be on your end."
>
> I understand where there are cases where the RTN message is sent because
> the call quality is actually terrible, but in those cases, there is usually
> several RTN messages and the sender will drop down to 4800 or below , and
> then usually give up and the call will fail.  This type of failure is rare
> (unless we have a major outage) and in this case the sending fax sees that
> it failed and will proceed with its normal retries.
>
> Question:
>
> Is this RTN to transmission failure hard coded, or is this a configurable
> setting?  If this were a regular fax machine, i think this would be a non
> issue as the receiver would see the crappy page as well as the good copy of
> that was sent again in their bundle of received pages.
>
> Any insight is greatly appreciated and for anyone just getting into FAX
> over IP with Cisco, I highly recommend the following book and Cisco Live
> presentations from these guys from Cisco TAC.
>
> Fax, Modem, and Text for IP Telephony [Book]
> <https://www.google.com/aclk?sa=l&ai=DChcSEwiXgOn0wY7TAhXJCioKHRVCANoYABAMGgJ0bQ&sig=AOD64_1ddnEIE3WNy8NH5g7yJ-WnrcChJA&ctype=5&q=&ved=0ahUKEwjei-P0wY7TAhUnj1QKHejtCy0QwzwIEg&adurl=>
> from Textbooks.com
> by David Hanes, Gonzalo Salgueiro
>
>
>
>
>
>
> Thanks,
>  Frank
>
>
>
> _______________________________________________
> cisco-voip mailing list
> cisco-voip at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-voip
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/cisco-voip/attachments/20170406/2d9ee500/attachment.html>