Recently we encountered an odd scenario with a Lync 2013 install:
- Dial a phone number from the Lync client
- Listen to 10 seconds of Lync’s ring tone (DING… ding)
- The call fails with “The call could not be completed…”
Testing the same phone number via a cell phone, we noted the following:
- About 11 – 13 seconds of silence (no ringing)
- Phone was then answered by an automated attendant
Using Lync Snooper, we noted the following in the logs.
- 5025;reason=”Cancel sent by application for INVITE client transaction.”;
NOTE: This error was generated 10 seconds after the initial SIP Invite - A SIP/2.0 487 Request Terminated was sent to the downstream VOIP provider
OK, it appears Lync simply hangs up the phone after 10 seconds if a response is not received by the called party.
Yes, we understand the called phone system was *very* slow, but to simply hang up is rather rude!
Resolution
What is really happing is Lync assumes the call failed and is trying to failover the outbound routing… and no failover exists (hence the hang-up)
To correct:
- Launch the Lync configuration web page and select Voice Routing
- Click Trunk Configuration, double click the Global record
- Uncheck Enable outbound routing failover timer
- Click OK and commit the changes
- Wait 5 – 10 minutes for the changes to take effect
Well done, but what are the consequences of disabling such safety mechanism? Maybe you caused a bigger issue than the other issue that you solved by disabling this.
The only consequence we can think of is if you actually have a failover route.
In which case, if you disable the failover timer, your failover routing will not be used.
But, if you actually had a failover route defined, this should never be an issue to begin with.
If every failover route takes more then 10 seconds to connect… say a call to Brazil, Africa, roaming cell phone etc… then LYNC will still drop the call.
What should be done is LYNC should have the ability to change this setting to larger then 10 seconds.
Thank you very much. I have applied your solution and it’s working fine.
You could manually increase the failover timer on the mediation servers themselves:
https://trogjels.wordpress.com/2012/03/22/outbound-call-from-lync-fails-timeout-issues/