Upstream health checks stall admin service after pod restarts in Kubernetes

Upstream health checks stall admin service after pod restarts in Kubernetes: a task in LegoFlow-SWE (Harbor dataset). No changes are required to the public API observable by the existing monitoring endpoints other than the addition of getPoolSize / setPoolSize for concurrency tuning.