Service-to-service authentication in Microservices - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding service-to-service authentication methods

Which of the following methods is most suitable for authenticating microservices communicating within a trusted internal network?

A Embedding static API keys in service requests without rotation

B Using mutual TLS (mTLS) certificates exchanged between services

C Relying solely on IP whitelisting without any cryptographic verification

D Using user session tokens passed between services

💡 Hint

Think about secure identity verification between services without relying on user context.

Mutual TLS (mTLS) provides strong cryptographic authentication between services by verifying certificates on both ends. Static API keys are less secure and lack rotation. IP whitelisting alone is weak and can be spoofed. User session tokens are meant for user-to-service authentication, not service-to-service.

❓ Architecture

intermediate

Designing a token-based service-to-service authentication flow

In a microservices architecture, which sequence correctly describes a typical OAuth 2.0 client credentials flow for service-to-service authentication?

A Service A requests an access token from the authorization server, then uses the token to call Service B

B Service A directly calls Service B without any token, relying on network security

C Service A sends user credentials to Service B to authenticate

D Service B requests a token from Service A before processing the request

💡 Hint

Consider which service obtains the token and who validates it.

In OAuth 2.0 client credentials flow, the calling service (Service A) obtains an access token from the authorization server and presents it to the called service (Service B) for authentication. Direct calls without tokens or user credentials are insecure or incorrect.

❓ scaling

advanced

Scaling service-to-service authentication with token caching

To reduce latency and load on the authorization server, a microservice caches access tokens. Which approach best balances security and performance?

A Request a new token for every service call without caching

B Cache tokens indefinitely without checking expiry to avoid repeated requests

C Cache tokens until they expire, then request new ones; validate token expiry before each use

D Cache tokens but ignore expiry and refresh only on failure

💡 Hint

Think about token validity and avoiding unnecessary calls.

Caching tokens until expiry reduces calls to the authorization server while ensuring tokens are valid. Ignoring expiry risks using invalid tokens. Requesting tokens every call increases latency and load.

❓ tradeoff

advanced

Choosing between mTLS and token-based authentication

Which statement best describes a tradeoff when choosing between mutual TLS (mTLS) and OAuth 2.0 token-based authentication for service-to-service security?

A mTLS provides strong identity verification but is complex to manage at scale; tokens are easier to rotate but require additional validation logic

B Tokens are more secure than mTLS because they use encryption; mTLS does not encrypt data

C Tokens cannot be revoked once issued; mTLS certificates never expire

D mTLS requires user credentials; tokens do not require any credentials

💡 Hint

Consider management complexity and security features.

mTLS offers strong cryptographic identity but managing certificates can be complex. Tokens simplify rotation and revocation but add validation overhead. The other options contain incorrect statements about security properties.

❓ estimation

expert

Estimating token validation load in a large microservices system

A system has 100 microservices, each making 50 authenticated calls per second to other services using OAuth 2.0 tokens. If each token validation takes 5 milliseconds on average, estimate the total CPU time spent per second on token validation across the system.

A 250 seconds of CPU time per second

B 2.5 seconds of CPU time per second

C 0.25 seconds of CPU time per second

D 25 seconds of CPU time per second

💡 Hint

Calculate total calls per second and multiply by validation time.

Total calls = 100 services * 50 calls = 5000 calls/sec. Each call validation takes 5 ms = 0.005 sec. Total CPU time = 5000 * 0.005 = 25 seconds per second. But since CPU time can be parallel, the question asks for total CPU time spent, which is 25 seconds. However, option D says 25 seconds, option D says 2.5 seconds. The correct calculation is 25 seconds, so option D is correct.