Identity and Authentication — System Design Academy

Complete System Design Tutorial

The design separates the latency-sensitive online path, the correctness-critical durable transition, and asynchronous projection or repair work. Each chapter below explains one decision shown in the diagrams.

1. Requirements and scope

Functional requirements

Execute the primary identity and authentication workflow and return an observable result.
Expose the concrete operations shown below with stable retry identities.
Allow operators to inspect, disable, repair, or replay ambiguous state.
Keep derived analytics and notifications off the correctness-critical path.

Non-functional requirements

Design for 1B identities · global login · p99 < 300 ms.
Protect the named business invariant under concurrency and partial failure.
Scale by an explicit aggregate/partition key without one global mutex.
Degrade intentionally when cache, queue, store, or regional ownership is unavailable.

Primary invariant: each accepted identity and authentication operation has one stable result and retries do not repeat its business effect.

↑ top

2. Capacity assumptions

State assumptions before choosing technology. The supplied workload is 1B identities · global login · p99 < 300 ms. Convert it into peak throughput, retained records, bandwidth, and failure budget; then connect each number to a component.

Dimension	Working estimate	Design implication
Traffic	1B identities · global login · p99 < 300 ms	Separate online compute, durable state, and asynchronous consumers.
Peak factor	5–10× average	Admission control, autoscaling headroom, and hot-key protection.
Stored state	write rate × retention × record bytes	Partition count, indexes, compaction, backup, and archival tier.
Network	peak operations × payload size	Batching, compression, edge placement, and egress budget.
Availability	99.9–99.99%	Multi-zone replicas, tested failover, and dependency error budgets.
Recovery	RPO/RTO by invariant	Ack boundary, replay retention, reconciliation, and operator runbooks.

Do not cargo-cult the numbers: show the arithmetic, name the uncertain variable, and say which production metric will replace the interview estimate.

↑ top

3. API and data model

POST /v1/loginprimary authentication

POST /v1/mfa/challenge

POST /v1/token/refresh

POST /v1/sessions/{id}/revoke

Authoritative records and indexes

identity(id PK, handles, status, credential_version)

credential(identity_id, type, verifier, parameters)

session(id, identity_id, device, expires_at, state)

token_key(kid, public_key, active_from, retired_at)

Command and event envelopes

# Command envelope
operation: POST /v1/login
request_id: req_01J...
idempotency_key: idem_7f2...
aggregate_key: identity-auth_123
expected_version: 17
payload: {...}

# Durable event
event_id: evt_01J...
aggregate_key: identity-auth_123
event_type: identity_auth_committed
schema_version: 3
trace_id: tr_8ac...

The authoritative aggregate and idempotency record share a transactional or conditional-write boundary. Event streams, search indexes, caches, analytics, and dashboards are rebuildable projections.

↑ top

4. Why this architecture and what it costs

The boxes are not a shopping list. Each boundary exists because it owns a different invariant, scaling dimension, data lifecycle, security policy, or failure domain. The table makes those choices explicit and shows when the design should remain simpler.

Component	Owned responsibility	Why separate it	Rejected alternative	Cost and evolution trigger
User / Workload	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Rate-limit login and verify credential in hardened boundary using current hash parameters.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents the external caller from also owning the contract and failure behavior of Auth Gateway.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Credential database is stolen” to the external caller.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the User / Workload boundary before extracting or scaling it.
Auth Gateway	Owns authentication, validation, quotas, deadlines, and routing. In this design it participates in: Evaluate risk/device context and require appropriate MFA or federation assertion.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents User / Workload from also owning the contract and failure behavior of Credential Verifier.	Put protocol, authentication, quotas, and business transitions in one process. Acceptable at low scale, but it couples external traffic risk to domain releases. Here that simpler choice would couple “Refresh token is replayed” to User / Workload.	Costs another hop, policy distribution, and a high-blast-radius fleet. Introduce it when many clients or services need one protocol and protection boundary. Measure the latency, saturation, lag, or mismatch at the Auth Gateway boundary before extracting or scaling it.
Credential Verifier	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Create server-side session and issue short-lived signed access token plus rotated refresh token.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Auth Gateway from also owning the contract and failure behavior of Password Hash Store.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Revocation stream lags” to Auth Gateway.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Credential Verifier boundary before extracting or scaling it.
Password Hash Store	Owns authoritative versioned state or immutable evidence. In this design it participates in: Services validate token locally; revocation and key updates stream to caches.	Defines the source of truth and commit boundary. It prevents Credential Verifier from also owning the contract and failure behavior of MFA Service.	Let each service persist its own interpretation. That removes one clear commit boundary and makes retries, support, and reconciliation ambiguous. Here that simpler choice would couple “Signing key must rotate immediately” to Credential Verifier.	Costs schema migration, replication, backup, and shard ownership. Split storage only when access patterns, retention, or compliance boundaries materially differ. Measure the latency, saturation, lag, or mismatch at the Password Hash Store boundary before extracting or scaling it.
MFA Service	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Recovery uses stronger proof, revokes existing sessions, and records immutable audit trail.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Password Hash Store from also owning the contract and failure behavior of Risk Engine.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Account recovery is socially engineered” to Password Hash Store.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the MFA Service boundary before extracting or scaling it.
Risk Engine	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Rate-limit login and verify credential in hardened boundary using current hash parameters.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents MFA Service from also owning the contract and failure behavior of Identity Directory.	Embed policy in every caller. Rules then drift by deployment version and emergency changes require fleet-wide redeployment. Here that simpler choice would couple “Credential database is stolen” to MFA Service.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Risk Engine boundary before extracting or scaling it.
Identity Directory	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Evaluate risk/device context and require appropriate MFA or federation assertion.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Risk Engine from also owning the contract and failure behavior of Session Store.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Refresh token is replayed” to Risk Engine.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Identity Directory boundary before extracting or scaling it.
Session Store	Owns authoritative versioned state or immutable evidence. In this design it participates in: Create server-side session and issue short-lived signed access token plus rotated refresh token.	Defines the source of truth and commit boundary. It prevents Identity Directory from also owning the contract and failure behavior of Token Signer / HSM.	Let each service persist its own interpretation. That removes one clear commit boundary and makes retries, support, and reconciliation ambiguous. Here that simpler choice would couple “Revocation stream lags” to Identity Directory.	Costs schema migration, replication, backup, and shard ownership. Split storage only when access patterns, retention, or compliance boundaries materially differ. Measure the latency, saturation, lag, or mismatch at the Session Store boundary before extracting or scaling it.
Token Signer / HSM	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Services validate token locally; revocation and key updates stream to caches.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Session Store from also owning the contract and failure behavior of Refresh Service.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Signing key must rotate immediately” to Session Store.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Token Signer / HSM boundary before extracting or scaling it.
Refresh Service	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Recovery uses stronger proof, revokes existing sessions, and records immutable audit trail.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Token Signer / HSM from also owning the contract and failure behavior of Revocation Stream.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Account recovery is socially engineered” to Token Signer / HSM.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Refresh Service boundary before extracting or scaling it.
Revocation Stream	Preserves durable ordering, replay, and consumer checkpoints. In this design it participates in: Rate-limit login and verify credential in hardened boundary using current hash parameters.	Decouples bursts and failure from user acknowledgment. It prevents Refresh Service from also owning the contract and failure behavior of Recovery Service.	Call every downstream consumer synchronously. That couples user latency and availability to analytics, notification, indexing, and audit dependencies. Here that simpler choice would couple “Credential database is stolen” to Refresh Service.	Costs eventual consistency, schema evolution, replay tooling, idempotent consumers, and lag operations. Earn it when bursts or downstream failure must not block acknowledgment. Measure the latency, saturation, lag, or mismatch at the Revocation Stream boundary before extracting or scaling it.
Recovery Service	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Evaluate risk/device context and require appropriate MFA or federation assertion.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Revocation Stream from also owning the contract and failure behavior of Federation / OIDC.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Refresh token is replayed” to Revocation Stream.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Recovery Service boundary before extracting or scaling it.
Federation / OIDC	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Create server-side session and issue short-lived signed access token plus rotated refresh token.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Recovery Service from also owning the contract and failure behavior of Device Trust.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Revocation stream lags” to Recovery Service.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Federation / OIDC boundary before extracting or scaling it.
Device Trust	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Services validate token locally; revocation and key updates stream to caches.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Federation / OIDC from also owning the contract and failure behavior of Audit / Abuse.	Keep this behavior inside an adjacent service. Prefer that simpler option until independent scale, ownership, policy, or failure isolation is measurable. Here that simpler choice would couple “Signing key must rotate immediately” to Federation / OIDC.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Device Trust boundary before extracting or scaling it.
Audit / Abuse	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Recovery uses stronger proof, revokes existing sessions, and records immutable audit trail.	Proves correctness and supports controlled recovery. It prevents Device Trust from also owning the contract and failure behavior of operators and repair tooling.	Infer correctness from request logs. Logs do not prove durable business state, detect silent drift, or safely repair ambiguous outcomes. Here that simpler choice would couple “Account recovery is socially engineered” to Device Trust.	Costs an interface, deployment, telemetry, ownership, and failure mode. Keep it in-process first; extract when independent scale or reliability pays for the boundary. Measure the latency, saturation, lag, or mismatch at the Audit / Abuse boundary before extracting or scaling it.

Decision 01

Chosen boundary vs simpler monolith

Passwords or keys are verified in a narrow hardened boundary; application services consume short-lived signed assertions rather than credentials.

A single deployable with one transactional store is the right starting point for a small identity and authentication workload. The proposed boundary becomes justified when passwords or keys are verified in a narrow hardened boundary; application services consume short-lived signed assertions rather than credentials. Separating it contains the named failure “Credential database is stolen” without placing unrelated work on the critical path.

Decision 02

Correctness vs availability

Short access tokens improve availability while refresh/session state enables revocation; fully stateless JWTs make immediate revocation expensive.

The design coordinates only where each accepted identity and authentication operation has one stable result and retries do not repeat its business effect. Everywhere else it accepts bounded staleness or asynchronous propagation. The alternative—strong coordination for every projection—simplifies the mental model but increases tail latency and turns partitions into broad write unavailability.

Decision 03

Operational cost and evolution

Risk and MFA are adaptive policy decisions with auditable versions, not hard-coded branches scattered across login clients.

This choice adds dashboards, ownership, schema compatibility, capacity planning, and recovery runbooks. Start with the minimum topology that supports “Create server-side session and issue short-lived signed access token plus rotated refresh token.” and extract another service only when measured throughput, blast radius, data lifecycle, or team ownership changes independently.

Why this design: the synchronous path contains only the decisions required to protect each accepted identity and authentication operation has one stable result and retries do not repeat its business effect. Durable projections, notifications, indexing, analytics, and repair are isolated so their load or failure cannot corrupt the acknowledged operation.

Correctness boundary

Passwords or keys are verified in a narrow hardened boundary; application services consume short-lived signed assertions rather than credentials.

Scale boundary

Short access tokens improve availability while refresh/session state enables revocation; fully stateless JWTs make immediate revocation expensive.

Operational boundary

Risk and MFA are adaptive policy decisions with auditable versions, not hard-coded branches scattered across login clients.

↑ top

5. Critical flow and algorithm

Rate-limit login and verify credential in hardened boundary using current hash parameters.

Evaluate risk/device context and require appropriate MFA or federation assertion.

Create server-side session and issue short-lived signed access token plus rotated refresh token.

Services validate token locally; revocation and key updates stream to caches.

Recovery uses stronger proof, revokes existing sessions, and records immutable audit trail.

Retry-safe mutation skeleton

identity = authenticate(request)
prior = idempotency.get(identity, request.key)
if prior: return prior.result

current = authoritative_store.read(partition_key)
next_state, events = domain.transition(current, request)
authoritative_store.compare_and_set(current.version, next_state)
idempotency.commit(request.key, next_state.result)
outbox.append(events)
return next_state.result

Exactly-once business effect: the network can duplicate requests. Stable identities, conditional state transitions, immutable event IDs, and reconciliation make the user-visible effect happen once.

↑ top

6. Caching, hot keys, and backpressure

Cache policy

Add a cache only for a measured repeated read; do not place mutable correctness state in a disposable layer.

Hot partition or tenant

Measure per-key and per-tenant skew. Replicate safe reads, isolate the offender, salt only when ordering permits, and enforce budgets.

Backpressure

Bound queues, buffers, connection pools, and concurrency. Shed low-priority work and scale from lag/oldest-item age rather than CPU alone.

Stampede and retry storms

Use single-flight/coalescing, jittered expiry, bounded retries, exponential backoff, and circuit breakers so one dependency failure does not multiply traffic.

Cache outage rule: never allow every miss to fall through simultaneously. Use local protection, request coalescing, circuit breakers, bounded pools, and controlled fallback.

↑ top

7. Partitioning and consistency

partition = stable_hash(primary_or_aggregate_key) → shard owner
write = conditional_update(expected_version)
read_projection = versioned and rebuildable

Strong boundary

each accepted identity and authentication operation has one stable result and retries do not repeat its business effect. Use a uniqueness constraint, serializable/conditional write, expected version, or fencing token at this boundary.

Eventual boundary

Analytics, notifications, search, read models, and operational dashboards may lag when freshness is observable and replay exists.

Operation	Consistency	Reason
Authoritative mutation	Conditional / serializable	Protect uniqueness, ownership, money, inventory, order, or legal state transition.
Read after own write	Primary, session token, or cache warm	The caller must observe the result it just committed.
Ordinary read projection	Bounded stale where safe	Replicas and caches improve latency when version/freshness is visible.
Analytics and operations	Eventual + replay	Freshness is less important than isolating the online path.

Route every mutation for one aggregate to one owner/partition epoch.
Use cursors rather than offsets for changing ordered collections.
Retain tombstones until replication and repair windows close.
Move shard ownership through versioned maps and dual-read verification.

↑ top

8. Failure drills and recovery

Failure	Runtime behavior	Recovery proof
Credential database is stolen	Stop unsafe progress, serve only explicitly safe degraded behavior, and recover from durable state.	Reconciliation returns zero invariant gaps; lag/replica health and user-visible SLO return to baseline.
Refresh token is replayed	Stop unsafe progress, serve only explicitly safe degraded behavior, and recover from durable state.	Reconciliation returns zero invariant gaps; lag/replica health and user-visible SLO return to baseline.
Revocation stream lags	Stop unsafe progress, serve only explicitly safe degraded behavior, and recover from durable state.	Reconciliation returns zero invariant gaps; lag/replica health and user-visible SLO return to baseline.
Signing key must rotate immediately	Stop unsafe progress, serve only explicitly safe degraded behavior, and recover from durable state.	Reconciliation returns zero invariant gaps; lag/replica health and user-visible SLO return to baseline.
Account recovery is socially engineered	Stop unsafe progress, serve only explicitly safe degraded behavior, and recover from durable state.	Reconciliation returns zero invariant gaps; lag/replica health and user-visible SLO return to baseline.

Recovery checkpoint

# Recovery checkpoint
scenario: Credential database is stolen
operation_id: op_01J...
aggregate_key: identity-auth_123
last_durable_version: 18
projection_checkpoint: partition-07@offset-482901
retry_policy: same_identity / bounded_backoff
reconciliation_status: pending
owner: Audit / Abuse

Every dependency needs a timeout, retry budget, circuit policy, and explicit fail-open, fail-closed, stale, or read-only mode. “Highly available” is incomplete until degraded behavior is named.

↑ top

9. Security, regions, and observability

Security and abuse

Authenticate at ingress and authorize the exact resource/tenant on every operation.
Encrypt transport and durable data; tokenize or isolate sensitive material.
Rate-limit by tenant, actor, resource, and network; detect enumeration and anomalous fan-out.
Redact secrets and personal data before logs or analytical storage.
Audit privileged mutations, support access, policy changes, and forced recovery.

Signals that prove correctness

User-visible success rate and p50/p95/p99 latency.
Timeouts, circuit state, saturation, and replica health by dependency.
Queue lag, oldest-item age, retries, and dead-letter growth.
Conflict, duplicate-suppression, and reconciliation mismatch counts.
Business drill: Credential database is stolen — alert before the user-visible invariant is violated.

Regional evolution

Begin multi-zone with tested failover, backups, and reconciliation.
Add global stateless ingress and regional read projections.
Keep one write owner per aggregate or partition where coordination is expensive.
Move ownership with epochs/fencing; avoid silent simultaneous writers.
Use expand/contract schemas, canaries, shadow reads, and reversible cutovers.

↑ top

10. Interviewer follow-ups

Session cookie or JWT?

Passwords or keys are verified in a narrow hardened boundary; application services consume short-lived signed assertions rather than credentials. Walk the concrete lifecycle step “Rate-limit login and verify credential in hardened boundary using current hash parameters.” Then test it against “Credential database is stolen”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How does token revocation work?

Short access tokens improve availability while refresh/session state enables revocation; fully stateless JWTs make immediate revocation expensive. Walk the concrete lifecycle step “Evaluate risk/device context and require appropriate MFA or federation assertion.” Then test it against “Refresh token is replayed”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How are passwords stored?

Risk and MFA are adaptive policy decisions with auditable versions, not hard-coded branches scattered across login clients. Walk the concrete lifecycle step “Create server-side session and issue short-lived signed access token plus rotated refresh token.” Then test it against “Revocation stream lags”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How does account recovery work?

Passwords or keys are verified in a narrow hardened boundary; application services consume short-lived signed assertions rather than credentials. Walk the concrete lifecycle step “Services validate token locally; revocation and key updates stream to caches.” Then test it against “Signing key must rotate immediately”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How do service identities differ?

Short access tokens improve availability while refresh/session state enables revocation; fully stateless JWTs make immediate revocation expensive. Walk the concrete lifecycle step “Recovery uses stronger proof, revokes existing sessions, and records immutable audit trail.” Then test it against “Account recovery is socially engineered”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

Answer pattern: state the invariant, choose a mechanism, walk one failure, name the recovery signal, and explain the cost of the alternative.

↑ top

11. How to present this in 45 minutes

Time	What to cover
0–5 min	Clarify scope, workload, SLOs, and the correctness invariant.
5–10 min	Estimate throughput/storage and define concrete APIs and records.
10–20 min	Draw online, durable-write, and asynchronous/repair paths.
20–30 min	Explain the exact lifecycle, partition key, and concurrency control.
30–38 min	Discuss caching, hot keys, replication, and backpressure.
38–43 min	Run failure drills, security, observability, and regional evolution.
43–45 min	Summarize the five most important choices and trade-offs.

Final decision summary

Invariant	each accepted identity and authentication operation has one stable result and retries do not repeat its business effect.
Partitioning	Stable aggregate/key hash with versioned ownership and measured hot-key isolation.
Commit	Expected version or uniqueness at the authoritative state boundary.
Async work	Durable outbox/log, idempotent consumers, checkpoints, and replay.
Recovery	Same operation identity, durable truth lookup, and business reconciliation.

Closing summary

CLOSING “The design protects each accepted identity and authentication operation has one stable result and retries do not repeat its business effect. The online path stays bounded, authoritative state is versioned and partitioned, and asynchronous work is replayable and observable.”

↑ top

User / Workload

Auth Gateway

Credential Verifier

Password Hash Store

Complete System Design Tutorial

1. Requirements and scope

Functional requirements

Non-functional requirements

2. Capacity assumptions

3. API and data model

Authoritative records and indexes

Command and event envelopes

4. Why this architecture and what it costs

Chosen boundary vs simpler monolith

Correctness vs availability

Operational cost and evolution

5. Critical flow and algorithm

Retry-safe mutation skeleton

6. Caching, hot keys, and backpressure

7. Partitioning and consistency

Strong boundary

Eventual boundary

8. Failure drills and recovery

Recovery checkpoint

9. Security, regions, and observability

Security and abuse

Signals that prove correctness

Regional evolution

10. Interviewer follow-ups

11. How to present this in 45 minutes

Final decision summary

Closing summary