Intermittent outbound attended transfer on Freeswitch

Hi Everyone,

I am running Freeswitch 1.10.8 on an AWS EC2 server.

I am not sure if this is a bug or a configuration issue on the Freeswitch because it is intermittent when we are doing attended transfers.

Here’s the scenario:
Person A calling Person B then Person B answers Person A
Person A calling Person C then Person C answers Person A
Person A initiate a transfer to connect Person B and Person C

Expectation: Person B should connect to Person C

The results are intermittent sometimes it is successful and sometimes failed.

When I tried to add a delay after the bridge let’s say sleep or delay for 10 seconds

The transfer is consistently successful but once the delay is done that’s the time the call drops.

That means, the attended transfer is successful but the Freeswitch tries to hangup the sessions which cause the attended transfer to fail.

I am rooting for a configuring issue at this moment but if there’s anything I missed in this scenario or configuration. That would be nice.

Parameters that I have played while debugging the feature:

  • hangup_after_bridge
  • continue_on_fail
  • park_after_bridge
  • park_timeout
  • transfer_after_bridge
  • ignore_early_media
  • absolute_codec_string
  • session setAutoHangup

I really appreciate the help.

Thank you.

Joenil

How are you doing this transfer? Via the buttons on the phone or in the dialplan with att_xfer?

Hi Brian,

The transfer is done via buttons on the phone.

What does the SIP trace show?

Hi Brian,

See the link below.
This is the SIP trace I got for failed attended transfer test.

Thank you so much. I appreciate the response.

Best regards,
Joenil

Are both call legs on the same FreeSWITCH Instance?

Hi Brian,

Yes, they are on the same FreeSWITCH instance.

I need the full freeswitch logs, with debug and sip traceon, fsctl loglevel debug, sofia loglevel all 9

/b

Hi Brian,

Here’s the link below of the freeswitch logs.

Thank you.

This level of interaction might require an FSE contract, I can’t clearly see why without getting into the system and man handling it while working with you, who would we speak with at your company about that?

/b

Are you doing anything with lua on these calls? Don’t set any variables to debug this, and do a new log without anything you’ve added.

/b

You’re clearing doing something with Lua, and not realizing it, the calls are being hungup in side some lua script switch_cpp.cpp:749 CoreSession::hangup