At 7:30 AM (Central), one of our load balancers failed. The redundant load balancer took over immediately, but a bug in the load balancer code prevented the redundancy from working properly. We had limited throughput for an extended period as we worked with the load balancer vendor to correct the issue. The extended downtime caused a large backlog of mail that took a while to process.
To prevent this issue going forward, we fixed the issue causing the load balancer to fail and tested the failover to ensure it is working properly.