July 2012 - FreeSWITCH-users - FreeSWITCH Mailing Lists

Fifo importance
by Chris B. Ware 30 Jul '12

30 Jul '12

Hi, Just another question on fifos (thanks for your answers on 'reparse'): what's the meaning of 'importance' parameter on fifo configuration? The same question is on wiki but with no answers. Thanks, Chris

1 0

Re: [Freeswitch-users] Push server and push capability
by Prasd D 29 Jul '12

29 Jul '12

Forgot ccs On 7/29/12, Prasd D <prasd.d.b(a)gmail.com> wrote: > Hi Regis, > > Thanks for throwing light on this. > > I fully agree with you on your first reason, and the initial general > comment. > >> AFAIK, what is done in C2DM is not magic. It should not be impossible to >> reach same thing on a third party app. > > I agree with you that we can have two sockets open. > > This is what I gather from what you put down above and have questions : > > - Presently push has only one socket open. 100% agree with you it > should not rely on a specific push provider > - A 3rd party app or something within csipsimple or a small component > of it will keep a socket open (while rest of csipsimple is shut off > for power purposes) through which it will receive a notification. This > notification will then start csipsimple to take an incoming call for > example. > Advantage of having a 3rd party app is that it can support push > notifications for other purposes also. However then again we are > relying on some other 3rd party push service provider. Instead if it > is done with the same SIP provider (achieved by using Freeswitch > hopefully it supports this as a module soon), then its even better. > I take it that this push service is then applicable only for this and > keeping a socket open just for this is not as optimal. > > > I didn't understand "solve the root cause on the app". I am not sure > if you are referring to "There is indeed known problems to keep the > connection alive on some devices with CSipSimple." when you are > talking about root cause. > But let's say csipsimple works fine when it is running. Then I guess > there is no "root cause" then? > > My main and only concern is saving power and battery and not having to > run csipsimple app running all the time just to receive incoming > calls. > So in this scenario (and no connection alive problems with > csipsimple), there is no "root cause" and the solution is just to use > push service. Is that correct ? > > Alternative is that csipsimple has components or plugins and while all > of it shuts off, one small component keeps running with a socket open > I guess to get simplied notification (rather than complex sip > messages) JUST to tell incoming call, whereafter the main csipsimple > starts. The Freeswitch server then waits for csipsimple to start and > send the invite (I am not sure how it will know that csipsimple woke > up, but I guess the job of the plugin is that. Perhaps it sens a push > notification and then few invites until it gets a response). > > Please clarify above. Would be great if you can provide a solution > together with freeswitch ! > > I think this is the future and as one of the developers indicated in > that jira bug, there are many companies now doing this (like > acrobits). > > > On 7/28/12, Régis Montoya <r3gis.3r(a)gmail.com> wrote: >> Hi Prasd, >> >> As I said many times to you about C2DM for invites... I still think >> that's a bad idea. >> >> There is indeed known problems to keep the connection alive on some >> devices with CSipSimple. >> But solving the issue by delegating to another solution hosted by google >> (or to any proprietary network) is not a good solution. If you want to >> develop something you should rather try to help on CSipSimple/pjsip for >> android to find out a reliable way for your devices. >> >> Several reason for that : >> * You become dependant of a push service provider... in this case >> google. Somebody using android should not be forced to use google. -and >> it also doesn't apply to private/enterprise networks-. >> * The C2DM doesn't ensure for reliability. Read the google docs, they >> tell that they don't assure that it will be delivered in time and for >> sure. It's the same for Apple push notification. >> * AFAIK, what is done in C2DM is not magic. It should not be >> impossible to reach same thing on a third party app. >> * Your sip provider becomes dependant of the push notification service >> provider : there is a mechanism to ensure your server is correctly >> allowed to publish on the push service provider (both for C2DM and Apple >> notifications). >> * Your application (in this case the CSipSimple plugin for your C2DM >> deployment), is also linked to the same mechanism and will be linked to >> certificates of the server (so also one app per service...) >> >> The only pro argument would be to have only one socket open, but in my >> opinion it's not a big deal. >> The other not receivable argument is laziness... (even if deployment of >> push notification can lead to headaches). >> >> If you really want to go this way feel free. The api provided by >> CSipSimple to other applications on android should allow you to do what >> you plan a clean and separated way. (I mean by a plugin). >> >> And I think that approach on FS is the same than mine : if you want such >> a thing, it should be addressed by a plugin. >> But my opinion remains the same, it's not a real solution to the >> problem. And I encourage any developer / sip provider facing this >> problem to help to focus their efforts to solve the root cause on the >> app ;). >> >> Best regards, >> Régis >> > > > -- > Thanks, > Prasd > -- Thanks, Prasd

1 0

Re: [Freeswitch-users] how to force sofia to use digest auth first?
by Mi Ke 29 Jul '12

29 Jul '12

I see... thank you very much for explaining ----- Original Message ----- From: Avi Marcus Sent: 07/29/12 08:39 PM To: FreeSWITCH Users Help Subject: Re: [Freeswitch-users] how to force sofia to use digest auth first? It's not what you think... part of the 401 unauthorized attempt is that it sends back a nonce key that you use to encrypt your password. That way, SIP has built-in protection against replay attacks even over unencrypted connections. (If someone is watching though, they can try to brute force your passwords based on the nonce key...) -Avi On Sun, Jul 29, 2012 at 8:05 PM, Mi Ke < mi.ke(a)null.net > wrote: Hi All, I use the following bridge params to originate call: {sip_auth_username=xxx,sip_auth_password=yyy,effective_caller_id_name=xxx,effective_caller_id_number=xxx}sofia/external/ 1111(a)1.1.1.1 Since my remote carrier supports only digest authentication, it replies with 401 (Unauthorized) for my first INVITE, and then FS falls back to digest auth and my calls goes OK. Is it possible to change auth type priority for sip profile/globally so digest auth will be used first or disable plain auth completely ? WBR / Mike _________________________________________________________________________ Professional FreeSWITCH Consulting Services: consulting(a)freeswitch.org http://www.freeswitchsolutions.com FreeSWITCH-powered IP PBX: The CudaTel Communication Server http://www.cudatel.com Official FreeSWITCH Sites http://www.freeswitch.org http://wiki.freeswitch.org http://www.cluecon.com Join Us At ClueCon - Aug 7-9, 2012 FreeSWITCH-users mailing list FreeSWITCH-users(a)lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org

1 0

how to force sofia to use digest auth first?
by Mi Ke 29 Jul '12

29 Jul '12

Hi All, I use the following bridge params to originate call: {sip_auth_username=xxx,sip_auth_password=yyy,effective_caller_id_name=xxx,effective_caller_id_number=xxx}sofia/external/1111(a)1.1.1.1 Since my remote carrier supports only digest authentication, it replies with 401 (Unauthorized) for my first INVITE, and then FS falls back to digest auth and my calls goes OK. Is it possible to change auth type priority for sip profile/globally so digest auth will be used first or disable plain auth completely ? WBR / Mike

2 1

ODBC, FreeTDS, Microsoft SQL
by Tim Meade 29 Jul '12

29 Jul '12

We wanted to run some ODBC tests using a MSSQL backend. We setup the ODBC with MySQL and everything worked great. Then installed FreeTDS and configured it instead of MySQL. When FS restarted, it created some of the required tables. Calls, complete, interfaces, nat, sip_presence, sip_recovery, sip_shared_apperance_subscriptions. What it did not create was any of the registrations tables. When a device tries to register, FS shows in the CLI: 012-07-28 09:43:48.892312 [INFO] switch_odbc.c:285 The connection has been re-established 2012-07-28 09:43:49.912319 [ERR] switch_odbc.c:494 ERR: [create index sr_sip_host on sip_registrations (sip_host)] [STATE: 42000 CODE 1088 ERROR: [unixODBC][FreeTDS][SQL Server]Cannot find the object "sip_registrations" because it does not exist or you do not have permissions. ] 2012-07-28 09:43:49.912319 [ERR] switch_core_sqldb.c:487 SQL ERR [STATE: 42000 CODE 1088 ERROR: [unixODBC][FreeTDS][SQL Server]Cannot find the object "sip_registrations" because it does not exist or you do not have permissions. ] create index sr_sip_host on sip_registrations (sip_host) 2012-07-28 09:43:49.912319 [CRIT] switch_odbc.c:280 The sql server is not responding for DSN freeswitch [STATE: 42000 CODE 1088 ERROR: [unixODBC][FreeTDS][SQL Server]Cannot find the object "sip_registrations" because it does not exist or you do not have permissions. ][244] Seems the issue is that FS didn't create all the appropriate tables on start. Has anyone else seen something like this? I cannot find much on using FS with FreeTDS and MSSQL. Thanks Tim

2 3

Google Inc. DOES use FreeSWITCH for call center telephony routing system. So, why dont' we invite those guys to FS weekly conference :)
by Anton Kvashenkin 29 Jul '12

29 Jul '12

Guys, according to those slides ( http://www.slideshare.net/OReillyOSCON/using-and-building-open-source-in-go… slide number 10) from Justin McWilliams ( https://plus.google.com/u/0/107742494065843953483/about ), Google does use FS for internal call center. So, I think that it would be great to invite some folks from Google to participate @ weekly FS conference. We can contact Justin via G+. For me it was really surprise, because I've never seen any guys from google committing something or discussing something @ ML. The agenda for that conference would be quite simple: to describe some use cases and the decision why did they choose to use FreeSWITCH.

6 11

Re: [Freeswitch-users] hunting in failover mode
by Mi Ke 28 Jul '12

28 Jul '12

<action application="set" data="fail_on_single_reject=USER_BUSY,NO_ANSWER"/> worked for me in combination with 1|2|3 construction ----- Original Message ----- From: Mi Ke Sent: 07/20/12 09:03 PM To: FreeSWITCH Users Help Subject: Re: [Freeswitch-users] hunting in failover mode Michael, thanks for answering - I already tried this but I need to set continue_on_fail to TRUE because that seems to be the only way to pass original hangup code from leg B to A when bridge completes. And when COF is set to TRUE it behaves like it's related to the "whole" bridge i.e. until all 3 endpoints are tried entire bridge is not considered as failed of succeded. Re separate approach - I'm not sure how can I do that as the number of EPs in my case is dynamic, bridge app receives them as a 1|2|3 dialstring from rouing lua app. Thanks / Mike ----- Original Message ----- From: Michael Collins Sent: 07/20/12 08:39 PM To: FreeSWITCH Users Help Subject: Re: [Freeswitch-users] hunting in failover mode Perhaps you could use this? http://wiki.freeswitch.org/wiki/Channel_Variables#continue_on_fail Specify the conditions on which the bridge should keep going. Note: you may need to separate these out into individual bridge apps with a single endpoint. I haven't tried it lately, so be sure to report back what you find out. -MC On Fri, Jul 20, 2012 at 10:30 AM, Mi Ke < mi.ke(a)null.net > wrote: Hi All, When I do a failover using bridge app and ep1|ep2|ep3 construstion - how do I stop further hunting if any of already hunted endpoints returned e.g. USER_BUSY ? Thanks / Mike _________________________________________________________________________ Professional FreeSWITCH Consulting Services: consulting(a)freeswitch.org http://www.freeswitchsolutions.com FreeSWITCH-powered IP PBX: The CudaTel Communication Server http://www.cudatel.com Official FreeSWITCH Sites http://www.freeswitch.org http://wiki.freeswitch.org http://www.cluecon.com Join Us At ClueCon - Aug 7-9, 2012 FreeSWITCH-users mailing list FreeSWITCH-users(a)lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org -- Michael S Collins Twitter: @mercutioviz http://www.FreeSWITCH.org http://www.ClueCon.com http://www.OSTAG.org

1 0

Push server and push capability
by Prasd D 28 Jul '12

28 Jul '12

I was searching and came across this amazing jira request for push server capability. This is exactly what I felt too, except didn't delve or earlier understand how to achieve this. http://jira.freeswitch.org/browse/FS-4347 This guy has identified very much what is really needed to be able to use it on smartphones. And suggests that Freeswitch and Csipsimple together show this capability where an incoming call or voicemail to a user using Csipsimple is notified through push notification. That starts Csipsimple (which is otherwise turned off) on his phone. This is a huge huge and amazingly useful thing since no one keeps their voip client on their phones on due to power consumption and battery drain ! This I think is the future of VOIP since smartphones are going to become soon huge % of voip clients.

3 4

Fifo reparse
by Chris B. Ware 28 Jul '12

28 Jul '12

Hi guys, if I have calls in fifos and try to add another fifo through XML curl using "fifo reparse", does it impact on active calls ? Thank you for your help, Chris

3 2

No reINVITE when bridging two sessions from mod_managed with bypass_media_after_bridge=true
by Srini K 28 Jul '12

28 Jul '12

Hi, Iam trying to bypass media from FS after two call legs are bridged using mod_managed. The code looks like public void Run(AppContext context) { var fsApi = new FreeSWITCH.Native.Api(); var aLegSession = context.Session; // Answer the incoming call aLegSession.Answer(); // Play the prompt aLegSession.StreamFile("ivr/ThankYou.wav", 0); // Create outBound session var bLegSession = new ManagedSession("sofia/gateway/95/4151230000"); // Bypass Media aLegSession.SetVariable("bypass_media_after_bridge", "true"); bLegSession.SetVariable("bypass_media_after_bridge", "true"); fsApi.ExecuteString(string.Format("uuid_bridge {0} {1}", aLegSession.GetUuid(), bLegSession.GetUuid())); } I don't see FreeSWITCH sending re-Invite after the call is bridged. What I've already tried and did not succeed: 1) set bypass_media=true, on A leg only, on B leg only, on both legs 2) set bypass_media_after_bridge=true, on A leg only, on B leg only, on both legs When I tried without using mod_managed using only dialplan, FS sends re-Invite. <action application="answer"/> <action application="sleep" data="1500"/> <action application="set" data="bypass_media_after_bridge=true"/> <action application="playback" data="ivr/ThankYou.wav"/> <action application="bridge" data="sofia/gateway/95/4151230000"/> Whether Iam doing anything stupid in mod_managed? Regards Srini

5 5