Service Issue in Voice Platform

Incident Report for Vonage API Status

Postmortem

Voice API Failures Due to High CPS and Invalid Traffic

 

Date/Time Impacted

Beginning - October 18th 2023, 19:14 UTC

Resolved - October 18th 2023, 20:26 UTC

‌

Services Impacted

Voice API

‌

Summary

Recently, we experienced a higher-than-usual volume of voice traffic, which led to a temporary disruption in our services. This was primarily due to an unexpected increase in call attempts per second (CPS) beyond the standard limits. Our systems, designed to ensure fair usage and optimal performance for all our users, automatically adjusted to manage this surge.

‌

Root Cause

During the impact window, there was an unexpected increase in call attempts per second (CPS) beyond the standard limits, some of which were invalid. 

Our system is designed to ensure fair usage and optimal performance for all our users. However, as it attempted to make automated adjustments to manage this surge, it resulted in a higher-than-expected number of calls being rejected, leading to a temporary service disruption for some of our customers.

‌

Restoration Actions

As a corrective measure, we implemented additional checks to block invalid calls and are improving our automated system's ability to adjust to these types of surges.

‌

Next Steps

Since restorative actions have been taken, all systems are stable.

Posted Nov 14, 2023 - 17:05 UTC

Resolved

This incident has been resolved.
Posted Oct 18, 2023 - 22:14 UTC

Monitoring

A fix has been implemented and we are monitoring the results.
Posted Oct 18, 2023 - 21:27 UTC

Investigating

Our monitoring has alerted us to a potential service issue causing intermittent connectivity issues to mainly US API endpoints. We are alerting customers immediately while our engineering teams investigate.

We aim to provide an update on our investigation within 1 hour.

Please see this article - https://api.support.vonage.com/hc/en-us/articles/6749696044828-Vonage-API-Support-Incident-Handling - for an explanation of our approach to publishing incidents.
Posted Oct 18, 2023 - 20:49 UTC
This incident affected: Multiple Services Affected (Multiple Services Affected) and Voice (Voice API).