Health checks and monitoring in Deno

Introduction

Health checks help us know if our app is working well. Monitoring keeps an eye on the app to catch problems early.

To check if a web server is running before sending users to it.

To automatically restart an app if it stops working.

To see how much memory or CPU the app is using.

To get alerts if the app is slow or down.

To keep a log of app health over time for improvements.

Syntax

Deno

import { serve } from "https://deno.land/std@0.203.0/http/server.ts";

serve((req) => {
  if (req.url.endsWith("/health")) {
    return new Response("OK", { status: 200 });
  }
  return new Response("Not Found", { status: 404 });
});

This example shows a simple HTTP health check endpoint.

The server responds with status 200 and message "OK" when the /health path is requested.

Examples

Basic health check endpoint returning 200 OK.

Deno

import { serve } from "https://deno.land/std@0.203.0/http/server.ts";

serve((req) => {
  if (req.url.endsWith("/health")) {
    return new Response("OK", { status: 200 });
  }
  return new Response("Not Found", { status: 404 });
});

Health check returns JSON with status and uptime info.

Deno

import { serve } from "https://deno.land/std@0.203.0/http/server.ts";

serve((req) => {
  if (req.url.endsWith("/health")) {
    const healthInfo = { status: "healthy", uptime: Deno.metrics().opsCompleted };
    return new Response(JSON.stringify(healthInfo), {
      status: 200,
      headers: { "Content-Type": "application/json" },
    });
  }
  return new Response("Not Found", { status: 404 });
});

Sample Program

This program starts a server that listens for HTTP requests. When you visit /health, it replies with "OK" and status 200, meaning the app is healthy. Other paths return 404 Not Found.

Deno

import { serve } from "https://deno.land/std@0.203.0/http/server.ts";

serve((req) => {
  if (req.url.endsWith("/health")) {
    return new Response("OK", { status: 200 });
  }
  return new Response("Not Found", { status: 404 });
});

Important Notes

Health check endpoints should be simple and fast to respond.

Use monitoring tools to regularly call the health check URL and alert if it fails.

Make sure health checks do not expose sensitive information.

Summary

Health checks tell if your app is running well.

Monitoring watches your app and alerts you to problems.

In Deno, you can create a simple HTTP endpoint that returns status 200 for health checks.

Practice

1. What is the main purpose of a health check endpoint in a Deno application?

easy

A. To verify if the application is running and responsive

B. To deploy the application to production

C. To increase the application's memory usage

D. To compile the Deno code into machine code

Solution

1. Step 1: Understand health check concept

   A health check endpoint is designed to confirm that the app is alive and responding to requests.

2. Step 2: Identify the correct purpose

   Options B, C, and D describe unrelated tasks like deployment, resource usage, or compilation, which are not health checks.

3. Final Answer:

   To verify if the application is running and responsive -> Option A

4. Quick Check:

   Health check = Verify app responsiveness [OK]

Hint: Health checks confirm app is alive and responding [OK]

Common Mistakes:

• Confusing health checks with deployment steps
• Thinking health checks improve performance directly
• Mixing health checks with code compilation

2. Which of the following is the correct way to create a simple health check HTTP endpoint in Deno?

easy

A. import { serve } from "https://deno.land/std/http/server.ts";
serve(req => console.log(req), { port: 8000 });

B. import { listenAndServe } from "https://deno.land/std/http/server.ts";
listenAndServe(8000, () => console.log("OK"));

C. import { serve } from "https://deno.land/std/http/server.ts";
serve(() => console.log("OK"), { port: 8000 });

D. import { serve } from "https://deno.land/std/http/server.ts";
serve(req => new Response("OK"), { port: 8000 });

Solution

1. Step 1: Check correct import and usage

   import { serve } from "https://deno.land/std/http/server.ts"; serve(req => new Response("OK"), { port: 8000 }); correctly imports 'serve' and returns a Response with "OK" for each request, which is a valid health check endpoint.

2. Step 2: Identify incorrect usage in other options

   import { listenAndServe } from "https://deno.land/std/http/server.ts"; listenAndServe(8000, () => console.log("OK")); uses deprecated 'listenAndServe' and does not respond to requests properly. Options A and C use console.log instead of returning a Response, so they won't respond correctly.

3. Final Answer:

   import { serve } from "https://deno.land/std/http/server.ts"; serve(req => new Response("OK"), { port: 8000 }); -> Option D

4. Quick Check:

   serve + Response("OK") = Correct health endpoint [OK]

Hint: Use serve with a Response object to reply to requests [OK]

Common Mistakes:

• Using console.log instead of returning Response
• Using deprecated functions like listenAndServe
• Not returning a response to HTTP requests

3. Given this Deno code snippet for a health check endpoint, what will be the output when accessed via HTTP?

import { serve } from "https://deno.land/std/http/server.ts";
serve(req => {
  if (new URL(req.url).pathname === "/health") {
    return new Response("Healthy", { status: 200 });
  } else {
    return new Response("Not Found", { status: 404 });
  }
}, { port: 8080 });

medium

A. Response with body 'Healthy' and status 404 when URL is '/health'

B. Response with body 'Healthy' and status 200 when URL is '/health'

C. Response with body 'Not Found' and status 200 for all URLs

D. Server crashes due to syntax error

Solution

1. Step 1: Analyze request URL condition

   The code checks if the request URL is exactly '/health'. If yes, it returns a 200 status with "Healthy".

2. Step 2: Understand else branch

   If the URL is anything else, it returns a 404 status with "Not Found".

3. Final Answer:

   Response with body 'Healthy' and status 200 when URL is '/health' -> Option B

4. Quick Check:

   URL '/health' = 200 Healthy [OK]

Hint: Check URL condition to predict response status and body [OK]

Common Mistakes:

• Confusing status codes for different URLs
• Assuming server crashes without syntax errors
• Mixing response body and status code

4. You wrote this Deno health check server code but it does not respond to requests properly. What is the error?

import { serve } from "https://deno.land/std/http/server.ts";
serve(req => {
  console.log("Received request");
});

medium

A. The serve function requires a port number as a string

B. The import statement is incorrect

C. The handler does not return a Response object

D. console.log is not allowed inside serve

Solution

1. Step 1: Check handler function

   The handler function only logs to console but does not return any Response, so HTTP requests get no valid reply.

2. Step 2: Verify other parts

   The import is correct, serve accepts an options object for port, and console.log is allowed. The main issue is missing Response return.

3. Final Answer:

   The handler does not return a Response object -> Option C

4. Quick Check:

   Handler must return Response [OK]

Hint: Always return a Response object in serve handler [OK]

Common Mistakes:

• Forgetting to return Response in handler
• Thinking console.log sends HTTP response
• Misunderstanding serve function parameters

5. You want to monitor your Deno app's health and expose metrics like uptime and request count. Which approach best combines health checks and monitoring in one server?

hard

A. Create two endpoints: '/health' returns 200 OK; '/metrics' returns JSON with uptime and request count

B. Use only '/health' endpoint returning plain text 'OK' without metrics

C. Create a single endpoint that returns HTML with uptime but no status code

D. Run two separate Deno servers, one for health and one for metrics, on different ports

Solution

1. Step 1: Understand combined health and monitoring needs

   Health checks confirm app is alive; monitoring provides metrics like uptime and request count.

2. Step 2: Evaluate options for best practice

   Create two endpoints: '/health' returns 200 OK; '/metrics' returns JSON with uptime and request count cleanly separates concerns with '/health' for status and '/metrics' for detailed data, both accessible on one server.

3. Step 3: Reject less effective options

   Use only '/health' endpoint returning plain text 'OK' without metrics lacks metrics, C lacks proper status code, D adds complexity by running two servers unnecessarily.

4. Final Answer:

   Create two endpoints: '/health' returns 200 OK; '/metrics' returns JSON with uptime and request count -> Option A

5. Quick Check:

   Separate endpoints for health and metrics = Best practice [OK]

Hint: Use separate endpoints for health and metrics on same server [OK]

Common Mistakes:

• Mixing metrics and health in one endpoint without status code
• Ignoring metrics in health checks
• Running multiple servers unnecessarily