Verify SNS messages delivered via HTTP(S) in Node.js

Michael Wittig – 14 Aug 2019

Are you implementing an HTTP/HTTPS endpoint for SNS? If so, you should definetly verify the incoming messages. Otherwise, anyone on the Internet can deliver messages to your HTTP/HTTPS endpoint. Which is a security risk.


How do you verify incoming messages? The SNS documentation answers this question:

You should verify the authenticity of a notification, subscription confirmation, or unsubscribe confirmation message sent by Amazon SNS.

In a nutshell, each SNS message contains a signature that we have to verify.

Verify SNS message: check the signature

The npm module sns-validator does the job. Unfortunately, the module is old and lacks support for save caching and certificate download retries. Therefore, I decided to implement this on my own, which wasn’t as hard as expected. Let’s get started.

First, you need to install a few dependencies:

  • request and requestretry to perform HTTP(S) requests with retries
  • lru-cache to safely cache certificates without running out of memory

Install the modules with:

npm i request requestretry lru-cache

Create a new JavaScript file (e.g., index.js) and import the dependencies we need:

const crypto = require('crypto');
const requestretry = require('requestretry');
const LRU = require('lru-cache');

According to the SNS documentation, we have to use different fields of the message based on the Type of the message delivered by SNS.

function fieldsForSignature(type) {
if (type === 'SubscriptionConfirmation' || type === 'UnsubscribeConfirmation') {
return ['Message', 'MessageId', 'SubscribeURL', 'Timestamp', 'Token', 'TopicArn', 'Type'];
} else if (type === 'Notification') {
return ['Message', 'MessageId', 'Subject', 'Timestamp', 'TopicArn', 'Type'];
} else {
return [];

We also have to come up with a way to download the certificate that we need to verify the signature. The certificate is attached to the message in the form of a URL. We have to download the certificate before we can verify the signature. Downloading things can fail for many reasons. Therefore, we retry failed download requests. To optimize for performance, we also want to cache downloaded certificates. Let’s look at the code.

const CERT_CACHE = new LRU({max: 5000, maxAge: 1000 * 60});

function fetchCert(certUrl, cb) {
const cachedCertificate = CERT_CACHE.get(certUrl);
if (cachedCertificate !== undefined) {
cb(null, cachedCertificate);
} else {
method: 'GET',
url: certUrl,
maxAttempts: 3,
retryDelay: 100,
timeout: 3000
}, (err, res, certificate) => {
if (err) {
} else {
if (res.statusCode === 200) {
CERT_CACHE.set(certUrl, certificate);
cb(null, certificate);
} else {
cb(new Error(`expected 200 status code, received: ${res.statusCode}`));

The cache stores a maximum of 5000 certificates and the certificates expire after 1 minute from the cache.

Last but not least, we do some input validation:

  • the fields SignatureVersion, SigningCertURL, Type, and Signature must be available
  • SignatureVersion must be 1
  • the SigningCertURL must start with https:// and we only want to download certificates from AWS
const CERT_URL_PATTERN = /^https:\/\/sns\.[a-zA-Z0-9-]{3,}\.amazonaws\.com(\.cn)?\/SimpleNotificationService-[a-zA-Z0-9]{32}\.pem$/;

function validate(message, cb) {
if (!('SignatureVersion' in message && 'SigningCertURL' in message && 'Type' in message && 'Signature' in message)) {
console.log('missing field');
cb(null, false);
} else if (message.SignatureVersion !== '1') {
console.log('invalid SignatureVersion');
cb(null, false);
} else if (!CERT_URL_PATTERN.test(message.SigningCertURL)) {
console.log('invalid certificate URL');
cb(null, false);
} else {
fetchCert(message.SigningCertURL, (err, certificate) => {
if (err) {
} else {
// TODO verifiy signature (insert next code block here)

Finally, the signature is verified.

const verify = crypto.createVerify('sha1WithRSAEncryption');
fieldsForSignature(message.Type).forEach(key => {
if (key in message) {
const result = verify.verify(certificate, message.Signature, 'base64');
cb(null, result);

You can test the code with a message like this:

Type: 'Notification',
MessageId: '4c807a89-9ef9-543b-bfab-2f4ed41e91b4',
TopicArn: 'arn:aws:sns:us-east-1:853553028582:marbot-dev-alert-Topic-8CT7ZJRNSA5Y',
Subject: 'INSUFFICIENT_DATA: "insufficient test" in US East (N. Virginia)',
Message: '{"AlarmName":"insufficient test","AlarmDescription":null,"AWSAccountId":"853553028582","NewStateValue":"INSUFFICIENT_DATA","NewStateReason":"tets","StateChangeTime":"2019-08-09T10:19:19.614+0000","Region":"US East (N. Virginia)","OldStateValue":"OK","Trigger":{"MetricName":"CallCount2","Namespace":"AWS/Usage","StatisticType":"Statistic","Statistic":"AVERAGE","Unit":null,"Dimensions":[{"value":"API","name":"Type"},{"value":"PutMetricData","name":"Resource"},{"value":"CloudWatch","name":"Service"},{"value":"None","name":"Class"}],"Period":300,"EvaluationPeriods":1,"ComparisonOperator":"GreaterThanThreshold","Threshold":1.0,"TreatMissingData":"- TreatMissingData: missing","EvaluateLowSampleCountPercentile":""}}',
Timestamp: '2019-08-09T10:19:19.644Z',
SignatureVersion: '1',
Signature: 'gnCKAUYX6YlBW3dkOmrSFvdB6r82Q2He+7uZV9072sdCP0DSaR46ka/4ymSdDfqilqxjJ9hajd9l7j8ZsL98vYdUbut/1IJ2hsuALF9nd/HwNLPPWvKXaK/Y3Hp57izOpeBAkuR6koitSbXX50lEj7FraaMVQfpexm01z7IUcx4vCCvZBTdQLbkWw+TYWkWNsMrqarW39zy474SmTBCSZlz1eoV6tCwYk2Z2G2awiXpnfsQRRZvHn4ot176oY+ADAFJ0sIa44effQXq+tAWE6/Z3M5rjtfg6OULDM+NGEmnVZL3xyWK8bIzB48ZclQo3ZsvLPGmCNQLlFpaP/3fGGg==',
SigningCertURL: '',
UnsubscribeURL: ''
}, (err, result) => {
if (err) {
} else {
console.log('result', result);


You should verify the authenticity of a message sent by Amazon SNS. The SNS documentation provides an in-depth description of the needed steps which can be implemented in Node.js as shown in this blog post.

Michael Wittig

Michael Wittig

I’m the author of Amazon Web Services in Action. I work as a software engineer, and independent consultant focused on AWS and DevOps.

You can contact me via Email, Twitter, and LinkedIn.

Briefcase icon
Hire me
Cover of Rapid Docker on AWS

New book: Rapid Docker on AWS

A rapid way to get your web application up and running on AWS. Made for web developers and DevOps engineers who want to dockerize their web applications and run their containers on Amazon Web Services. Prior knowledge of Docker and AWS is not required.

Buy icon
Buy now
Marbot Logo

Incident Management for Slack

Team up to solve incidents with our chatbot marbot. Never miss a critical alert. Escalate alerts from your AWS infrastructure among your team members. Strong integrations with all parts of your AWS infrastructure: CloudWatch, Elastic Beanstalk, RDS, EC2, ...

Slack icon
Try for free
🎉 Rapid Docker on AWS out now
Learn to bundle your application into Docker containers and run them on AWS. Get the eBook or join the Online Seminar.