Honest to Death

A load balancer health check hits an endpoint that runs a real query against the shared database. The database slows down for 40 seconds, every instance fails its check, and the balancer removes the entire fleet, turning a partial degradation into a total outage. What is the right fix?

Provision additional app servers so that a larger number of them stay healthy while the database is degraded

Shorten the check interval from 10 seconds to 2 so failing hosts leave rotation faster

Raise the consecutive-failure threshold from two checks to three so brief stalls are tolerated

Use a shallow liveness check and let the balancer fail open once too few hosts remain healthy