Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-26 Thread Mark Michelson
Brian J. Murrell wrote:
> On Fri, 2008-09-26 at 10:16 -0400, SIP wrote:
>> The RFCs are there for a reason. All SIP forking is UAS territory. Not
>> UAC territory.
> 
> From http://bugzilla.gnome.org/show_bug.cgi?id=553810 Damien Sandras
> asks:
> 
> I repeat, Ekiga is doing something perfectly legal.
> 
> The real question is why does Asterisk think it is the same request 
> when the
> from tag is different ?
> 
> b.
> 

Asterisk ignores tags in To and From headers unless you have "pedantic=yes" set 
in sip.conf.

Mark Michelson

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-26 Thread Brian J. Murrell
On Fri, 2008-09-26 at 10:41 -0400, SIP wrote:
> Oh yes. It's perfectly legal.
> 
> It's also a) NOT SIP forking, b) Lazy, and c) Poorly designed.
> 
> Sending multiple requests and hoping and praying that the recipient will
> ignore two of them (it will NOT in many cases -- specifically set out by
> the RFC -- see MESSAGE) because the tag is different doesn't make it any
> less poorly designed just because it's not specifically written that it
> can't be done.

The Ekiga developer points out:

Have a look at this link :
http://www.faqs.org/rfcs/rfc3261.html

And look how an UAS (Asterisk in this case) is supposed to handle merged
requests :
8.2.2.2 Merged Requests

   If the request has no tag in the To header field, the UAS core MUST
   check the request against ongoing transactions.  If the From tag,
   Call-ID, and CSeq exactly match those associated with an ongoing
   transaction, but the request does not match that transaction (based
   on the matching rules in Section 17.2.3), the UAS core SHOULD
   generate a 482 (Loop Detected) response and pass it to the server
   transaction.

  The same request has arrived at the UAS more than once, following
  different paths, most likely due to forking.  The UAS processes
  the first such request received and responds with a 482 (Loop
  Detected) to the rest of them.

In our case, the From tag is different, so it should not detect a loop.

This excerpt show that any client or server should be able to receive 
merged
requests. There is obviously a bug here in Asterisk.

I have no idea how (in-)correct it is, again, just being the messenger.

b.



signature.asc
Description: This is a digitally signed message part
___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users

Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-26 Thread SIP
Brian J. Murrell wrote:
> On Fri, 2008-09-26 at 10:16 -0400, SIP wrote:
>   
>> The RFCs are there for a reason. All SIP forking is UAS territory. Not
>> UAC territory.
>> 
>
> From http://bugzilla.gnome.org/show_bug.cgi?id=553810 Damien Sandras
> asks:
>
> I repeat, Ekiga is doing something perfectly legal.
> 
> The real question is why does Asterisk think it is the same request 
> when the
> from tag is different ?
>
> b.
>
>   
Oh yes. It's perfectly legal.

It's also a) NOT SIP forking, b) Lazy, and c) Poorly designed.

Sending multiple requests and hoping and praying that the recipient will
ignore two of them (it will NOT in many cases -- specifically set out by
the RFC -- see MESSAGE) because the tag is different doesn't make it any
less poorly designed just because it's not specifically written that it
can't be done.

N.

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-26 Thread Brian J. Murrell
On Fri, 2008-09-26 at 10:16 -0400, SIP wrote:
> 
> The RFCs are there for a reason. All SIP forking is UAS territory. Not
> UAC territory.

From http://bugzilla.gnome.org/show_bug.cgi?id=553810 Damien Sandras
asks:

I repeat, Ekiga is doing something perfectly legal.

The real question is why does Asterisk think it is the same request 
when the
from tag is different ?

b.



signature.asc
Description: This is a digitally signed message part
___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users

Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-26 Thread Watkins, Bradley


> -Original Message-
> From: [EMAIL PROTECTED] 
> [mailto:[EMAIL PROTECTED] On Behalf Of 
> Brian J. Murrell
> Sent: Friday, September 26, 2008 10:12 AM
> To: Asterisk Users Mailing List - Non-Commercial Discussion
> Subject: Re: [asterisk-users] sip forking needed for ekiga 3.0

> 
> > I've read enough of the thread to know the Asterisk issue you are
> > trying to describe is loop detection and not forking. Asterisk does
> > support forking: Dial(SIP/user1&SIP/user2) is forking. Not 
> being able
> > to handle duplicate requests from different IPs is loop handling and
> > you'll already find bugs open about that.
> 
> I will relay your description of "loop detection" back on to the Ekiga
> guys.  I'm just the messenger here.


Can you provide a SIP trace of the signalling taking place?  What are
you getting back from the Asterisk box?

- Brad

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-26 Thread SIP
A machine with more than one default gateway is a VERY special case
(used for load-balancing or possibly failover). Most systems will not
allow it. I mean... logically, it's odd. Default means "when not applied
to any other special rule, I choose this one."Not this two. Not this
three. This one. It used to be called the Gateway of Last Resort. Last
being final and not penultimate.

With that being said, if you somehow manage to get by the internal
consistency checks and more than one interface (and by interface, I also
mean alias, as those are 'virtual interfaces') matches the default
gateway, your machine is misconfigured and internet traffic will not
properly flow.

I know you're just the messenger here, and it's not your fault. But the
message is wrong. Ekiga has tried to solve a problem (that of
determining a 'best path' for SIP to allow data flow in a NAT or
filtered scenario) using poorly thought-out logic. While there may be
any number of SIP proxies out there (SER is one of them, and I know
that's what the Ekiga service uses) that might be able to handle a
mistake on the client side with ease and grace, there's no guarantee
that they all will, and assuming they will simply because your test
environment allows it is lazy.

The RFCs are there for a reason. All SIP forking is UAS territory. Not
UAC territory.

N.

Brian J. Murrell wrote:
> On Fri, 2008-09-26 at 08:43 +0100, Grey Man wrote:
>   
>> It's not particularly difficult to determine the best IP address for a
>> piece of client software to use.
>> 
>
> Oh?
>
>   
>> Check the local machines default
>> gateway, apply the subnet mask and then compare it against all the
>> local IP's.
>> 
>
> Yeah?  And if more than one matches?  Then what?
>
> Have you read the whole thread here?
>
> b.
>
>   
> 
>
> ___
> -- Bandwidth and Colocation Provided by http://www.api-digital.com --
>
> AstriCon 2008 - September 22 - 25 Phoenix, Arizona
> Register Now: http://www.astricon.net
>
> asterisk-users mailing list
> To UNSUBSCRIBE or update options visit:
>http://lists.digium.com/mailman/listinfo/asterisk-users


___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-26 Thread Brian J. Murrell
On Fri, 2008-09-26 at 14:54 +0100, Grey Man wrote:
> On Fri, Sep 26, 2008 at 2:36 PM, Brian J. Murrell <[EMAIL PROTECTED]> wrote:
> >
> > Yeah?  And if more than one matches?  Then what?
> >
> 
> Use one of them!

And if the one I "choose" to use doesn't work because of some kind of
policy routing or filtering, etc.?

> 
> And if the network set up is too complex that that still causes
> problems do what the Ekiga guy told you and set the IP address you
> want to use in the config file.

Uhm.  That's not an option.  There is no option to do that.  Should
there be?  Perhaps.

But hey, I'm just here to report the "failure" (i.e. to properly accept
multiple "legal" registrations through this so called SIP forking) that
the Ekiga guys are telling me about.

> I've read enough of the thread to know the Asterisk issue you are
> trying to describe is loop detection and not forking. Asterisk does
> support forking: Dial(SIP/user1&SIP/user2) is forking. Not being able
> to handle duplicate requests from different IPs is loop handling and
> you'll already find bugs open about that.

I will relay your description of "loop detection" back on to the Ekiga
guys.  I'm just the messenger here.

b.



signature.asc
Description: This is a digitally signed message part
___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users

Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-26 Thread Grey Man
On Fri, Sep 26, 2008 at 2:36 PM, Brian J. Murrell <[EMAIL PROTECTED]> wrote:
>> Check the local machines default
>> gateway, apply the subnet mask and then compare it against all the
>> local IP's.
>
> Yeah?  And if more than one matches?  Then what?
>

Use one of them!

And if the network set up is too complex that that still causes
problems do what the Ekiga guy told you and set the IP address you
want to use in the config file. It's not a difficult situation. I have
servers with 15 public IP addresses on them and manage to run SIP
services no problems.

I've read enough of the thread to know the Asterisk issue you are
trying to describe is loop detection and not forking. Asterisk does
support forking: Dial(SIP/user1&SIP/user2) is forking. Not being able
to handle duplicate requests from different IPs is loop handling and
you'll already find bugs open about that.

> Oh?

I like to make sure I've done my homework before ascending to sarcasm
especially when I'm the one that requested "Thots" in the first
place...

Greyman.

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-26 Thread Brian J. Murrell
On Fri, 2008-09-26 at 08:43 +0100, Grey Man wrote:
> 
> It's not particularly difficult to determine the best IP address for a
> piece of client software to use.

Oh?

> Check the local machines default
> gateway, apply the subnet mask and then compare it against all the
> local IP's.

Yeah?  And if more than one matches?  Then what?

Have you read the whole thread here?

b.



signature.asc
Description: This is a digitally signed message part
___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users

Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-26 Thread Grey Man
"Shotgunning" the use of IP addresses is foolish at best and lazy
programming at worst. Imagine if the poeple writing browsers did that!
The internet could end up with double or triple the traffic for no
extra benefit not to mention the additinoal load on web servers etc.

It's not particularly difficult to determine the best IP address for a
piece of client software to use. Check the local machines default
gateway, apply the subnet mask and then compare it against all the
local IP's.

Regards,

Greyman.

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-25 Thread SIP
Alex Balashov wrote:
> You need to define what you mean by "SIP forking."  There are many 
> things people mean by that.  They are usually one of:
>
> 1) Call branching (proxies do this).
>
> 2) Parallel but distinct call legs managed by a UAC (this is what 
> Asterisk does when you Dial(SIP/exten1&SIP/exten2&SIP/exten3,...)).
>
>   
Exactly. These are all endpoint or middlepoint things. SIP forking is 
never an original starting point thing. That's just WEIRD. You fork to 
hit multiple endpoints simultaneously. Not one endpoint from multiple 
starting points.

N.

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-25 Thread Alex Balashov
You need to define what you mean by "SIP forking."  There are many 
things people mean by that.  They are usually one of:

1) Call branching (proxies do this).

2) Parallel but distinct call legs managed by a UAC (this is what 
Asterisk does when you Dial(SIP/exten1&SIP/exten2&SIP/exten3,...)).

-- 
Alex Balashov
Evariste Systems
Web: http://www.evaristesys.com/
Tel: (+1) (678) 954-0670
Direct : (+1) (678) 954-0671
Mobile : (+1) (706) 338-8599

___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-25 Thread Brian J. Murrell
On Thu, 2008-09-25 at 15:31 -0400, SIP wrote:
> That strikes me as being careless and unreliable.

That's one argument.  I can also see the ekiga developers' argument
though and that's to strive for the most automatic functionality
possible.  The less things you have to ask users, the more likely you
are to "just work".

> Call me a purist, but 
> I'm of the opinion that you should KNOW which interface to use based on 
> which interface is registered

We are talking about IP aliases here, not real interfaces.

> and choose ONE interface based on the 
> rules you've established during registration.

What rules would you establish during registration?

> What happens if you want 
> to ensure that data goes across a VPN (in order to encrypt your VoIP 
> communications) instead of the public internet?

Presumably you have some [policy] routing that ensures that.  But on the
other hand, if you did have two addresses on an interface, one for VPN
and one for "everything else", unless you "shotgun" out you need to
either know which address to use or ask the user.  Either case may fail.

> That takes all the logic out of the equation and just says, "Here's a 
> bunch of packets. Figure out what to do with them. I'll be waiting for 
> your response."

I don't think it's quite that bad.  It's more like here's a bunch of
session requests, please complete them [you don't know it yet, but I'm
going to tear down all but the first one you complete].  But the glitch
is that even though I send you 3 of them, due to [policy] routing and
firewalling, you might only get one.

> There's a reason routing rules exist and mature services allow you to 
> control the interface from which it originates.

Really, I'm just the messenger here.  I doubt the ekiga team and the
asterisk team would be willing to sit down and discuss who is right
here, so I'm trying to be the conduit.

b.



signature.asc
Description: This is a digitally signed message part
___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users

Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-25 Thread SIP
That strikes me as being careless and unreliable. Call me a purist, but 
I'm of the opinion that you should KNOW which interface to use based on 
which interface is registered and choose ONE interface based on the 
rules you've established during registration. What happens if you want 
to ensure that data goes across a VPN (in order to encrypt your VoIP 
communications) instead of the public internet? Or if you want to ensure 
a particular route based on why you created your multiple interfaces in 
the first place?

That takes all the logic out of the equation and just says, "Here's a 
bunch of packets. Figure out what to do with them. I'll be waiting for 
your response."

There's a reason routing rules exist and mature services allow you to 
control the interface from which it originates.

N.


Brian J. Murrell wrote:
> On Thu, 2008-09-25 at 14:56 -0400, SIP wrote:
>   
>> Sending from multiple different points of origin doesn't make any sense
>> at all in either a logical or rational fashion. What's it supposed to
>> accomplish?
>> 
>
> It seems to be a "shot-gun" approach to making a SIP connection.  The
> assumption being I suppose that one or more of the IP aliases will fail
> for whatever reason (policy routing, filtering, etc.), so just try them
> all, and use the first one to make a completion and drop the others.
>
> b.
>
>   
> 
>
> ___
> -- Bandwidth and Colocation Provided by http://www.api-digital.com --
>
> AstriCon 2008 - September 22 - 25 Phoenix, Arizona
> Register Now: http://www.astricon.net
>
> asterisk-users mailing list
> To UNSUBSCRIBE or update options visit:
>http://lists.digium.com/mailman/listinfo/asterisk-users


___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users


Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-25 Thread Brian J. Murrell
On Thu, 2008-09-25 at 14:56 -0400, SIP wrote:
> Sending from multiple different points of origin doesn't make any sense
> at all in either a logical or rational fashion. What's it supposed to
> accomplish?

It seems to be a "shot-gun" approach to making a SIP connection.  The
assumption being I suppose that one or more of the IP aliases will fail
for whatever reason (policy routing, filtering, etc.), so just try them
all, and use the first one to make a completion and drop the others.

b.



signature.asc
Description: This is a digitally signed message part
___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users

Re: [asterisk-users] sip forking needed for ekiga 3.0

2008-09-25 Thread SIP
My thoughts are that to do parallel requests from every IP address on
the machine is extremely weird behaviour.

How would any server know which to respond to?

SIP forking is supposed to send requests to multiple different
destinations (or fork mid-stream to send to different destinations). 
Sending from multiple different points of origin doesn't make any sense
at all in either a logical or rational fashion. What's it supposed to
accomplish?

N.

Brian J. Murrell wrote:
> So, I have been testing ekiga 3.0 with Asterisk, and sadly, it don't
> work.  I am told by the ekiga devs in
> http://bugzilla.gnome.org/show_bug.cgi?id=553595 and
> http://bugzilla.gnome.org/show_bug.cgi?id=553810 that the problem is
> that Asterisk does not support SIP forking.
>
> The issue is that I have multiple addresses on my workstation:
>
> 2: eth0:  mtu 1500 qdisc pfifo_fast qlen 1000
> link/ether xx:xx:xx:xx:xx:xx brd ff:ff:ff:ff:ff:ff
> inet 10.75.22.1/24 brd 10.75.22.255 scope global eth0
> inet 10.75.22.101/24 brd 10.75.22.255 scope global secondary eth0:1
>
> So when ekiga (3.0) tries to place a call through Asterisk it in fact
> does parallel requests from all addresses.  This is what appears to
> confuse Asterisk.  Please see the above tickets for more details.
>
> Thots?
>
> b.
>
>   
> 
>
> ___
> -- Bandwidth and Colocation Provided by http://www.api-digital.com --
>
> AstriCon 2008 - September 22 - 25 Phoenix, Arizona
> Register Now: http://www.astricon.net
>
> asterisk-users mailing list
> To UNSUBSCRIBE or update options visit:
>http://lists.digium.com/mailman/listinfo/asterisk-users


___
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

AstriCon 2008 - September 22 - 25 Phoenix, Arizona
Register Now: http://www.astricon.net

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users