Hi Bogdan,
We upgraded to v0.8.14, and t_check_status works as you described, and perfectly for what we want to accomplish.
Thank you,
Dave
----- Original Message -----
From: Bogdan-Andrei IANCU <iancu(a)fokus.fraunhofer.de>
Date: Thursday, July 29, 2004 3:19 am
Subject: Re: [Serusers] t_check_status/t_on_failure does not work with ser v.0.8.12 rpm
> David Lee wrote:
> > Hello,
> >
> > In the admin documentation of ser, it says that all messages
> greater than 300 are considered failures when using t_on_failure.
> >
> > In our scenario, we would like the ability to forward to a voice
> mail on 4xx failure, but not on 302 failure. We are using XTEN
> Pro softphones and want to be able to allow users the
> functionality to forward to another SIP url (uses 302). But when
> ser is seeing 302 (and t_on_failure is active) it considers it a
> failure and immediately runs failure route block (to voicemail).
> So our users cannot using forwarding to URL.
> >
> > An example would be ... When I'm out for lunch I would like to
> forward my line to my secretary first before going to voicemail.
> >
> > We saw some mention of t_check_status ... but ser wont start
> with that in the cfg. We are using ser v.0.8.12 rpm. Is that in
> a unstable version only. If some could post how to accomplish
> this it would be appreciated.
>
> yes, indeed, the answer is t_on_failure - this function didn't
> existed
> in the original release 0.8.12 (so, not present in rpm), but you
> can
> find it in the cvs tree of 0.8.12 - so, get the sources from cvs.
>
> bogdan
>
Kenny Lam wrote:
>Hi bogdan,
>
> I mean the calls are cumulative calls.
>
cumulative calls doesn't load ser. from ser point of view, the call is
processed only when it's initialised and terminated. So, basically you
can have as much active call as you want, without loading ser at all.
anyhow, it's an impresiv number of parallel call. It's a stress test or
....? if it not, may I ask the number of registered user?
> I received the following error log
>when it was down.
>
> 0(4798) ERROR: build_res_buf_from_sip_req: out of memory ; needs 584
> 0(4798) ERROR: build_res_buf_from_sip_req: out of memory ; needs 584
> 0(4798) ERROR: sl_reply_error used: I'm terribly sorry, server error occured
>(1/SL)
> 0(4798) ERROR: build_res_buf_from_sip_req: out of memory ; needs 533
> 0(4798) ERROR: build_req_buf_from_sip_req: out of memory
> 0(4798) ERROR: print_uac_request: no pkg_mem
> 0(4798) ERROR: t_forward_nonack: failure to add branches
> 0(4798) ERROR: build_res_buf_from_sip_req: out of memory ; needs 584
> 0(4798) ERROR: build_res_buf_from_sip_req: out of memory ; needs 584
> 0(4798) ERROR: sl_reply_error used: I'm terribly sorry, server error occured
>(1/SL)
> 2(4805) ERROR: fifo_server fgets failed: Illegal seek
> 2(4805) ERROR: fifo_server fgets failed: Illegal seek
> 2(4805) ERROR: fifo_server fgets failed: Illegal seek
> 2(4805) ERROR: fifo_server fgets failed: Illegal seek
>
>
>
looks like your ser gets into troubles because of insufficient private
memory (default is 1M). Try to increase it in config.h,
PKG_MEM_POOL_SIZE value (it's a define).
about the crash...excepting this log, do you have any additional info
that will help us to track the problem down? no coredump is generated?
bogdan
>Thanks a lot for your kindly help!
>
>
>On Fri, 30 Jul 2004 10:18:38 +0200, Bogdan-Andrei IANCU wrote
>
>
>>Hi Kenny,
>>
>>first what do you mean by 12000-15000 calls? established calls in
>>parallel or calls that initiated in the same time? there is a huge
>>diference.
>>second....what kind of failures? can you be a little bit more
>>specific
>>(even c'n'p from log will do it) third... when ser goes down.. there
>>is no core dumped? no message in the log related to receiving a signal?..
>>
>>regards,
>>bogdan
>>
>>Kenny Lam wrote:
>>
>>
>>
>>>Hi all,
>>>
>>> I have setup the SER with media proxy to deal with the problem of NAT.
>>>However, each time the number of calls reaches around 12000-15000, the SER
>>>will either doesn't accept call and increase failurerates or it will be down
>>>without any prior warnings. Anyone got the same experience before? I've tried
>>>using both CVS 8.12-1rc6 and rc5 before, both get the same problem.
>>>
>>>Thanks a lot!
>>>Best Regards,
>>>
>>>Kenny Lam
>>>SIP Application Engineer
>>>Deltapath Commerce & Technology Limited
>>>---------------------------------------
>>>SIP By Deltapath!
>>>www.deltapath.com
>>>
>>>_______________________________________________
>>>Serusers mailing list
>>>serusers(a)lists.iptel.org
>>>http://lists.iptel.org/mailman/listinfo/serusers
>>>
>>>
>>>
>>>
>>>
>
>
>Kenny Lam
>SIP Application Engineer
>Deltapath Commerce & Technology Limited
>---------------------------------------
>SIP By Deltapath!
>www.deltapath.com
>
>
>
>
I'm trying out this what looks like nice software while
the kids are sleeping.
I just installed 0.8.14 and I'm messing around with getting
multiple domains to play. What is the correct way to update
the table domain ? Just an INSERT directly into the database ?
Regards Tony
--
---
Tony Sarendal - tony.sarendal(a)polarcap.org
Cisco/Unix/Babies
-= The scorpion replied,
"I couldn't help it, it's my nature." =-
Totally agree with Jiri. An ALG unnecessarily brings up the price of the ADSL modem as well.
---
Dhiraj Bhuyan
Network Security Specialist,
BT Exact Business Assurance Solutions
Tel: +44 1473 643932
Mob: +44 7962 012145
Email: dhiraj.2.bhuyan(a)bt.com
-----Original Message-----
From: serusers-bounces(a)iptel.org [mailto:serusers-bounces@lists.iptel.org]On
Behalf Of Jiri Kuthan
Sent: 30 July 2004 10:46
To: andres(a)telesip.net; serusers(a)lists.iptel.org
Subject: Re: [Serusers] Broken "SIP Aware" Router
At 11:40 PM 7/29/2004, Andres wrote:
>We have started having problems with customers in BellSouths ADSL network. New customers are being given the Westell ADSL modem which appears to be "SIP Aware". But the implementation is terribly broken.
Well, I think that Application-Level-Gateways (ALGs) is _fundamentally_ broken
concept with very negative impact on interoperability, security and software
engineering.
There will be attempt to create an effort documenting these issues in the upcoming
IETF meeting; till this effort brings fruits it will take some time. Some enlighted
vendors are aware of this: Cisco/Linksys gave up on the idea to introduce SIP
awareness in residential routers.
Thank you for your report.
-jiri
>The symptoms are that SIP INVITES cannot penetrate the NAT on the Modem even though we have a permanent Keep-Alive from the inside (every 15 seconds). Furthermore when the customer tries to make a call, the modem changes the source port of every single packet of the dialog. Thus breaking the SIP dialog completely.
>
>Our interim solution was to setup these customer on an alternate server listening on a different port.
>For everybody's reference the modem version is: Westell B9061003006 and the Software is: 03.00.61
>
>I have tried to contact Westell directly but they have not answered. I hope this is not something thats going to start happening with all modem manufacturers or else were going to have loads of trouble.
>
>
>--
>Andres
>Network Admin
>http://www.telesip.net
>
>
>_______________________________________________
>Serusers mailing list
>serusers(a)lists.iptel.org
>http://lists.iptel.org/mailman/listinfo/serusers
--
Jiri Kuthan http://iptel.org/~jiri/
_______________________________________________
Serusers mailing list
serusers(a)lists.iptel.org
http://lists.iptel.org/mailman/listinfo/serusers
hello, I just start learned sip and sems, can you tell me how to use
sems+voicemail?
can you give me the ser.cfg file and tell me how to test voicemail? thanks
Sean
We have started having problems with customers in BellSouths ADSL
network. New customers are being given the Westell ADSL modem which
appears to be "SIP Aware". But the implementation is terribly broken.
The symptoms are that SIP INVITES cannot penetrate the NAT on the Modem
even though we have a permanent Keep-Alive from the inside (every 15
seconds). Furthermore when the customer tries to make a call, the modem
changes the source port of every single packet of the dialog. Thus
breaking the SIP dialog completely.
Our interim solution was to setup these customer on an alternate server
listening on a different port.
For everybody's reference the modem version is: Westell B9061003006 and
the Software is: 03.00.61
I have tried to contact Westell directly but they have not answered. I
hope this is not something thats going to start happening with all modem
manufacturers or else were going to have loads of trouble.
--
Andres
Network Admin
http://www.telesip.net
Can someone please help me figure out why I am getting this error from my SER:
ERROR: t_newtran: EoH not parsed
The message it is responding to is this:
ACK sip:user@172.18.1.100:5060 SIP/2.0
Via: SIP/2.0/UDP 172.18.1.45:5060
From: "Cory Zue" <sip:czue@172.18.1.100>;tag=2005707378
To: <sip:user@172.18.1.100>;tag=1047340964
Call-ID: 1555446853(a)172.18.1.45
CSeq: 101 ACK
User-Agent: Cory's VoIP UA
Route: <sip:user@172.18.1.35:5060>
Content-Length: 0
Every line in the message ends with a "\r\n".
Any help with this problem, or where to look to get more output (see what
line causes the error) would be greatly appreciated.
thanks in advance,
Cory
Hi all,
I have setup the SER with media proxy to deal with the problem of NAT.
However, each time the number of calls reaches around 12000-15000, the SER
will either doesn't accept call and increase failurerates or it will be down
without any prior warnings. Anyone got the same experience before? I've tried
using both CVS 8.12-1rc6 and rc5 before, both get the same problem.
Thanks a lot!
Best Regards,
Kenny Lam
SIP Application Engineer
Deltapath Commerce & Technology Limited
---------------------------------------
SIP By Deltapath!
www.deltapath.com
