Dominique –
I was considering the second option you
mentioned, we have a very similar application using a perl api.
If you check often enough and resend faxes
that failed, it should accomplish the behavior you described. It would however
have to resend any fax that gets a busy signal, or any other kind of
communications failure. In our environment, we send a lot of faxes to providers
that get busy signals, so this would cause a whole lot of churn, faxes failing,
getting deleted and getting resent. This looks like too much of a liability on
our side.
Lee – I would appreciate any insight
on the jobcontrol option for this issue, knowing that it is unsupported, as a
temporary workaround until the feature is added in a new version. What are the
potential downsides of resetting the “start page counter” ?
-Siri Vias
From: Dominique
dHotman [mailto:dhotman@xxxxxxxxx]
Sent: Wednesday, November 12, 2008
11:21 AM
To: Siri
Vias Khalsa
Cc: lee.howard@xxxxxxxxxxxx;
hylafax-users@xxxxxxxxxxx
Subject: Re: [hylafax-users]
Continuation on failure...
Thank you very much for the detail on this, I was quite uncertain about
the systems behavior and am grateful for the clarification.
This is unfortunately a problem for our users which I need to
find a work around for. Ideally we would be able to configure the option
to use continuation or not but as this isn't currently available I'll have to
make another plan.
The two options that I can see right now:
- Do something with with jobcontrol that resets the
number of pages sent and in so doing resends all pages of the fax instead
of sending from the last confirmed page. I'm not sure where to even
start with this so any input would be greatly appreciated.
- The way that our environment is setup at the
moment is that we have some custom built middleware that sits between our
business applications and in front of the HylaFAX server. This
middleware takes care of things like document conversion, cover page
creation, and various other bits of business logic. The middleware
uses the GNU HylaFAX Java API to create fax jobs and has the ability to
set different job parameters (because of the API); one of these parameters
is retries. So what we could try doing is make the retry count 1
when we create the fax job and then have the middleware handle retries
instead of HylaFAX handling them. This way to HylaFAX it would be a
new job every time and would therefore try send the entire fax and not
just from the last confirmed page. Any view on whether this would
work or not?
Last thing, how would I go about initiation some sort of enhancement
request to have the configuration option for continuation added?
On Wed, Nov 12, 2008 at 7:29 PM, Siri Vias
Khalsa <sirivias.khalsa@xxxxxxxxxx>
wrote:
Thanks Lee.
Dominic - I believe this answers your question.
-----Original Message-----
From: Lee Howard [mailto:lee.howard@xxxxxxxxxxxx]
Sent: Tuesday, November 11, 2008 10:57 AM
To: Siri Vias Khalsa
Subject: RE: [hylafax-users] Continuation on failure...
Siri Vias,
The behavior of HylaFAX has always (since at least 1999) been the same:
if
it has an error when it retries the fax it will start at the first page
that
was not confirmed by the receiver as received. So if you have a 50-page
fax
that fails on page 10 (the last page confirmed was page 9) then on the
next
retry it will begin at page 10.
This behavior is not configurable in the usual way. I suspect that
jobcontrol could be used to modify the sendq file to alter npages or
something. But it would be an undocumented or untested approach.
If the behavior is desired I think that the right way to do it is to
have a
typical configuration feature for it. It probably wouldn't be difficult
to
implement.
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*