The outbound queue is throwing SIP/488 Not Acceptable Here the second the pacing multiplier pushes past 1.2x. It’s only happening on the premium contact list that’s flagged for DNC exemption. Admin UI shows the SIP trunk status as completely green, latency sitting at 45ms, but the disposition log floods with SIP_FAILURE_488 right before the call connects. Pacing gets forced down to 0.9x automatically, which tanks the hourly dial rate. System is running the 2024-11-19 release on a US-East tenant. Callable windows are set to 09:00-17:00 ET, and the campaign’s abandonment threshold is locked at 3.0%.
Checked the SIP trunk configuration in the admin portal. Codec negotiation is set to G.711u, and the invite timeout is 30s. Nothing’s changed on the carrier side. The weird part is the error only triggers when the predictive engine batches more than 15 simultaneous invites. Drop the batch size and the trunks swallow the calls fine. Compliance team is already breathing down my neck because the pacing drop is messing with the scheduled wrap-up times.
Console logs on the campaign dashboard show predictive_engine: throttle_triggered alongside the 488 codes. Tried toggling the Progressive fallback mode, but that just queues everything and kills throughput. The SIP dialog trace shows the INVITE gets sent, the carrier replies with 100 Trying, then immediately slams back the 488. No 408 or 503 in sight.
Looking at the raw SIP headers in the trace, the P-Asserted-Identity header is pulling the wrong outbound caller ID from the contact list override. Might be why the carrier’s rejecting it. Admin UI isn’t giving a way to force a specific CLI per batch. Just sitting at 0.9x pacing while the queue backs up.