Module 15 · API Security in Microservices Mesh

Manish Garg
Manish Garg Associate of (ISC)² · RingSafe
Apr 27, 2026
2 min read
Read as

Last updated: April 29, 2026

100% Free

No signup. No paywall. No catch. One of our 10 most-requested practitioner modules — published in full so anyone can learn for free. We earn through consulting, not by gating knowledge.

See all 10 free modules →

Why this module. Most API-security advice covers north-south (internet to API). In microservices, east-west traffic (service to service) is 10x more volume and often less protected. Compromise one service, lateral movement to others.

Why this module. Most API-security advice covers north-south (internet to API). In microservices, east-west traffic (service to service) is 10x more volume and often less protected. Compromise one service, lateral movement to others.

The trust model that fails

“Internal services trust each other; auth happens at the edge.” Once an attacker is inside (via vuln in any service), they can hit every other service freely. Modern threat model: zero trust east-west too.

What service mesh provides

Module 11 (Cloud track) covers Istio / Linkerd in depth. For APIs:

  • mTLS automatic — every service-to-service call encrypted + authenticated
  • Identity per service — “service A” is a verifiable cryptographic identity
  • Policy enforcement — service A can call service B on /v1/users only
  • Observability — every cross-service call logged with identity

Identity propagation

User authenticates at edge → token issued. Token reaches downstream services. Each service must verify the token (don’t blindly trust upstream).

Pattern:

  • Edge gateway validates user JWT
  • Gateway issues an internal token signed by internal CA, with user identity embedded
  • Internal services verify internal token; act with user identity
  • Audit log captures user + service chain

Authorization in microservices

Two patterns:

  • Centralised — every service calls a Policy Decision Point (PDP) for authz. Open Policy Agent (OPA) is the standard.
  • Distributed — each service has authz code. Faster but inconsistent.

Centralised is harder to scale; distributed has consistency drift. Modern pattern: OPA sidecar per service for low latency + central policy.

Audit logging across services

Trace ID propagates through every service. Audit log captures:

  • Trace ID
  • User identity
  • Service chain (A → B → C)
  • Resource accessed
  • Decision (allowed / denied)

Forensics: query “what did user X do” returns the full cross-service chain.

Defender’s checklist

  • Service mesh deployed; mTLS strict.
  • Identity propagation pattern (edge token + internal token).
  • Per-service authorization policy.
  • OPA or equivalent for centralised policy.
  • Distributed tracing with trace ID through all services.
  • Audit log captures cross-service flows.
🧠
Check your understanding

Module Quiz · 6 questions

Pass with 80%+ to mark this module complete. Unlimited retries. Each question shows an explanation.

Want this for your team?

Custom team training + practitioner advisory

Beyond the free academy — we run private workshops, vCISO advisory, and red-team exercises tailored to your stack. For Indian SMBs scaling past their first hire.

Book team training call Replies in 4 working hrs · India-only · Senior consultants