But here's the terrifying part. Because wap-03 was "alive" according to basic ICMP pings, the cluster's consensus protocol had been treating it as a voting member. For six months, every time wap-03 choked on a null byte, it would delay the cluster's session replication by 400ms.

I pulled the plug on wap-03 at 2:53 AM.

For six months, wap-03 had been a source of low-grade anxiety. Every Tuesday at 4:00 PM, latency on that node would spike by 200ms. The logs showed a cryptic error: Event ID 1309 – Connection dropped by backend . Management refused to let me take it offline. "It's redundant," my boss, Linda, had said. "Redundancy means we keep it."