Distributes incoming traffic
Backend & Runtime / Knowledge library
Health Check
Tests whether an instance is alive, ready, or functioning
Why learn Health Check?
It guides routing, restart, and deployment decisions
The problem it solves
It keeps traffic away from instances unable to serve correctly
Where it came from
Health Check is a networking and web concept that tests whether an instance is alive, ready, or functioning. Its guarantees and tradeoffs depend on the concrete implementation and operating context.
Build your understanding
Work through these exercises in order. Try the task before opening a reference.
- Define Health Check and distinguish it from a neighboring concept.
- Choose a realistic task where Health Check is appropriate.
- Identify a failure mode caused by misusing Health Check.
- Interpret a small design or result involving Health Check.
- Explain one tradeoff or limitation of Health Check.
Documentation and sources
- MDN Web Technology Glossary (developer.mozilla.org)
Connect the ideas
Explore next
Real World Example
Automated Resilience for Distributed Order Processing
A regional logistics firm managing a distributed order processing system that relies on microservices to track inventory and shipping status.
The business problem
When a specific service instance hangs due to memory pressure or deadlocked threads, the load balancer continues routing traffic to the unresponsive node, causing cascading request timeouts and customer-facing errors.
How this topic solves it
Implementing a health check mechanism allows the load balancer to periodically poll a dedicated endpoint on each service instance. If an instance fails to return a successful status code, the load balancer automatically removes it from the rotation, ensuring traffic only reaches functional nodes.
- Define a lightweight HTTP endpoint on each service instance dedicated to reporting status.
- Configure the load balancer to perform periodic GET requests against this specific health endpoint.
- Set a threshold for consecutive failed checks before marking the instance as unhealthy.
- Implement automated removal logic to eject unhealthy instances from the active traffic pool.
- Enable automatic re-insertion once the instance returns a successful status code after recovery.
This diagram illustrates how a load balancer monitors service instances to manage traffic flow.
Active processing node
Unresponsive node
Status reporting interface
- Load Balancer → Routes Traffic → Service Instance A
- Load Balancer → Polls Status → Health Endpoint
- Health Endpoint → Reports Failure → Service Instance B
- Load Balancer → Ejects Instance → Service Instance B
Expected business outcome
The system achieves self-healing capabilities where traffic is dynamically rerouted away from failing components, significantly reducing the duration of service disruptions for end users.
Limitations and checks
A significant tradeoff is the risk of false negatives where a service is marked unhealthy due to transient network jitter rather than actual application failure. To verify this, monitor the frequency of state transitions in the load balancer logs to ensure they align with actual service crashes rather than temporary network spikes.