The incident occurred when a Google Cloud TCP load balancer in the us-west1-c region failed, causing around 5% of Netlify's global traffic to be disrupted. The issue was due to a bug in Google's networking agent that removed a custom route table entry, preventing the backend nodes from responding with the correct IP address. To mitigate the incident, Netlify implemented alternative IPs for enterprise and self-serve customers, and eventually restored all traffic after restarting the GCP networking daemon on affected nodes. The resolution of the incident was delayed due to a simultaneous cloud provider outage in the same region. Netlify is now patching the guest-agent systemd config to listen to systemd-networkd restarts and expanding its load balancer monitoring to include backend health checks. Additionally, improvements are planned to be rolled out this year to further improve reliability, scalability, and mean time to resolution.