Call/session count suddenly spiking up and reaching limit

We’re running 6 identical instances that handle traffic evenly distributed among them and 1 of the instances suddenly hit the maximum session limit.

We’re logging the call count and session count every minute and it looks like within a period of 30 minutes the call and session count went from around 100 to 1000 (current limit defined) and this shows on the logs:

2023-05-04 14:41:21.578434 71.57% [CRIT] switch_core_session.c:2432 Over Session Limit! 1000

We haven’t seen anything unusual on the logs and would be helpful of anyone can share tips on how to go about this. We’re wondering if outbound Calls (leg b) are not getting cleared, even after call gets hang-up · Issue #960 · signalwire/freeswitch · GitHub can be related (although we’re not using early media).

Thank you

What revision of FreeSWITCH are you running?


Are you connected to ESL doing anything specific?


We basically control the entire call flow via ESL yes. We invoke ASR and TTS on every interaction to communicate.

Can you give an example of what you do? Sounds like you’re blocking the socket some how and causing this.