Skip to main content

​
Overview

The HTTP Load Balancer uses structured logging to provide visibility into all operations. The Logger class (src/utils/logger.ts) categorizes log output into five types: REQUEST, RESPONSE, HEALTH, ERROR, and INFO.

​
Log categories

​
REQUEST logs

Logs every incoming request and which backend server it’s routed to. Format: REQUEST: {method} {path} -> {backendUrl}
src/utils/logger.ts
static request(method: string, path: string, backendUrl: string): void {
  console.log(`REQUEST: ${method} ${path} -> ${backendUrl}`);
}
Example output: 
REQUEST: GET / -> http://localhost:3001
REQUEST: POST /api/users -> http://localhost:3002
REQUEST: GET /health -> http://localhost:3003
Request logs are emitted in src/proxy/proxyHandler.ts:23 immediately after a backend is selected.

​
RESPONSE logs

Logs the response from the backend including status code and response time. Format: RESPONSE: {method} {path} <- {backendUrl} [{statusCode}] {duration}ms
src/utils/logger.ts
static response(method: string, path: string, backendUrl: string, statusCode: number, duration: number): void {
  console.log(`RESPONSE: ${method} ${path} <- ${backendUrl} [${statusCode}] ${duration}ms`);
}
Example output: 
RESPONSE: GET / <- http://localhost:3001 [200] 12ms
RESPONSE: POST /api/users <- http://localhost:3002 [201] 45ms
RESPONSE: GET /health <- http://localhost:3003 [200] 8ms
Use response duration metrics to identify slow backends. Consistently high response times may indicate backend performance issues.

​
HEALTH logs

Logs the health status of each backend server during periodic health checks. Format: HEALTH: {status} {url}{extraInfo}
src/utils/logger.ts
static health(url: string, isHealthy: boolean, details?: string): void {
  const status = isHealthy ? 'HEALTHY' : 'UNHEALTHY';
  const extraInfo = details ? ` - ${details}` : '';
  console.log(`HEALTH: ${status} ${url}${extraInfo}`);
}
Example output: 
HEALTH: HEALTHY http://localhost:3001 - status: 200
HEALTH: HEALTHY http://localhost:3002 - status: 200
HEALTH: UNHEALTHY http://localhost:3003 - Connection refused
HEALTH: UNHEALTHY http://localhost:3004 - status: 500

​
ERROR logs

Logs error conditions including proxy failures and missing backends. Format: ERROR: {message}{backend}
src/utils/logger.ts
static error(message: string, backendUrl?: string): void {
  const backend = backendUrl ? ` (${backendUrl})` : '';
  console.error(`ERROR: ${message}${backend}`);
}
Example output: 
ERROR: No healthy backends available
ERROR: Backend failed after 2034ms (http://localhost:3002)
ERROR logs indicate service degradation. “No healthy backends available” means the load balancer cannot serve any requests.

​
INFO logs

Logs general informational messages about load balancer state and operations. Format: INFO: {message}
src/utils/logger.ts
static info(message: string): void {
  console.log(`INFO: ${message}`);
}
Example output: 
INFO: Load balancer running on port 3000
INFO: Backend servers: http://localhost:3001, http://localhost:3002, http://localhost:3003
INFO: Health checker started (checking every 5s)

​
Sample log output

Here’s a complete log sequence showing startup, health checks, and request handling: 
INFO: Load balancer running on port 3000
INFO: Backend servers: http://localhost:3001, http://localhost:3002, http://localhost:3003
INFO: Health checker started (checking every 5s)
HEALTH: HEALTHY http://localhost:3001 - status: 200
HEALTH: HEALTHY http://localhost:3002 - status: 200
HEALTH: UNHEALTHY http://localhost:3003 - Connection refused
REQUEST: GET / -> http://localhost:3001
RESPONSE: GET / <- http://localhost:3001 [200] 12ms
REQUEST: GET / -> http://localhost:3002
RESPONSE: GET / <- http://localhost:3002 [200] 8ms
REQUEST: GET / -> http://localhost:3001
RESPONSE: GET / <- http://localhost:3001 [200] 10ms

​
Monitoring backend health

​
Health check flow

1

Parallel health checks

The health checker queries all backends concurrently using Promise.all() every 5 seconds (configurable).
2

Timeout enforcement

Each health check request has a 3-second timeout enforced via AbortController. Slow backends are marked unhealthy.
3

Status evaluation

Backends returning HTTP 2xx status codes are marked healthy. All other responses (4xx, 5xx) or connection failures mark the backend as unhealthy.
4

Automatic recovery

Once a previously unhealthy backend starts responding successfully, it’s automatically re-added to the healthy pool.

​
Tracking backend status

You can monitor backend health by watching HEALTH log lines:
  • HEALTHY with status code → Backend is operational
  • UNHEALTHY with error message → Backend is down or misconfigured
  • UNHEALTHY with non-2xx status → Backend is reachable but returning errors

​
Log aggregation

All logs are written to stdout/stderr. For production deployments, pipe logs to a log aggregation service like Datadog, Splunk, or the ELK stack.
Example with Docker: 
docker logs -f <container-id> | tee loadbalancer.log
Example with systemd: 
journalctl -u loadbalancer -f

Build docs developers (and LLMs) love

Get started for freeTalk to us