Unique ID Generator — 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 unique id generator 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 10M IDs/s · multi-region · 64–128 bits.
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 unique id generator 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 10M IDs/s · multi-region · 64–128 bits. Convert it into peak throughput, retained records, bandwidth, and failure budget; then connect each number to a component.

Dimension	Working estimate	Design implication
Traffic	10M IDs/s · multi-region · 64–128 bits	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

GET /v1/ids?count=allocate batch

POST /internal/workers/leaseunique worker identity

POST /internal/ranges/leasenumeric range

GET /v1/ids/{id}/decodetimestamp/region diagnostics

Authoritative records and indexes

worker_lease(worker_id PK, owner, generation, lease_until)

range_lease(start, end, owner, generation, consumed)

epoch(version, start_time, bit_layout)

collision(id, first_owner, second_owner, observed_at)

Command and event envelopes

# Command envelope
operation: GET /v1/ids?count=
request_id: req_01J...
idempotency_key: idem_7f2...
aggregate_key: unique-id-generator_123
expected_version: 17
payload: {...}

# Durable event
event_id: evt_01J...
aggregate_key: unique-id-generator_123
event_type: unique_id_generator_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
Service Client	Owns client protocol, request identity, retry, and resume behavior. In this design it participates in: Acquire a unique worker/region identity through consensus and persist generation locally.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents the external caller from also owning the contract and failure behavior of Regional Endpoint.	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 “Clock moves backward” 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 Service Client boundary before extracting or scaling it.
Regional Endpoint	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Read monotonic time; detect wall-clock rollback before composing an ID.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Service Client from also owning the contract and failure behavior of ID Worker.	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 “Two processes share worker ID” to Service Client.	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 Regional Endpoint boundary before extracting or scaling it.
ID Worker	Executes idempotent asynchronous work under lease and retry policy. In this design it participates in: Pack timestamp, region, worker, and per-millisecond sequence into configured bits.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Regional Endpoint from also owning the contract and failure behavior of Monotonic Clock.	Perform the work before acknowledging the caller. This gives simpler consistency but unbounded tail latency and fragile retries. Here that simpler choice would couple “Sequence overflows one millisecond” to Regional Endpoint.	Costs leases, checkpoints, retries, DLQs, and lag operations. Separate once execution time or burstiness no longer fits the request deadline. Measure the latency, saturation, lag, or mismatch at the ID Worker boundary before extracting or scaling it.
Monotonic Clock	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: When sequence exhausts, wait for next tick or consume a leased numeric range.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents ID Worker from also owning the contract and failure behavior of Worker Lease Cache.	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 “Epoch/bit layout changes during rollout” to ID Worker.	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 Monotonic Clock boundary before extracting or scaling it.
Worker Lease Cache	Serves a low-latency, disposable projection with authoritative fallback. In this design it participates in: Stop generation on lost lease and monitor sampled/full uniqueness at system boundaries.	Removes repeated work from the latency-critical path. It prevents Monotonic Clock from also owning the contract and failure behavior of Consensus Lease Service.	Read authoritative storage for every request. Simpler initially, but latency and source-of-truth load scale together and a hot key becomes a database incident. Here that simpler choice would couple “Identifiers leak business volume” to Monotonic Clock.	Costs invalidation, staleness, stampede protection, and memory. Add or split it only after the workload (10M IDs/s · multi-region · 64–128 bits) proves repeated reads dominate. Measure the latency, saturation, lag, or mismatch at the Worker Lease Cache boundary before extracting or scaling it.
Consensus Lease Service	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Acquire a unique worker/region identity through consensus and persist generation locally.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Worker Lease Cache from also owning the contract and failure behavior of Worker Registry.	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 “Clock moves backward” to Worker Lease Cache.	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 Consensus Lease Service boundary before extracting or scaling it.
Worker Registry	Executes idempotent asynchronous work under lease and retry policy. In this design it participates in: Read monotonic time; detect wall-clock rollback before composing an ID.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Consensus Lease Service from also owning the contract and failure behavior of Epoch Config.	Perform the work before acknowledging the caller. This gives simpler consistency but unbounded tail latency and fragile retries. Here that simpler choice would couple “Two processes share worker ID” to Consensus Lease Service.	Costs leases, checkpoints, retries, DLQs, and lag operations. Separate once execution time or burstiness no longer fits the request deadline. Measure the latency, saturation, lag, or mismatch at the Worker Registry boundary before extracting or scaling it.
Epoch Config	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Pack timestamp, region, worker, and per-millisecond sequence into configured bits.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Worker Registry from also owning the contract and failure behavior of Rollback Detector.	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 “Sequence overflows one millisecond” to Worker Registry.	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 Epoch Config boundary before extracting or scaling it.
Rollback Detector	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: When sequence exhausts, wait for next tick or consume a leased numeric range.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Epoch Config from also owning the contract and failure behavior of Range Allocator.	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 “Epoch/bit layout changes during rollout” to Epoch Config.	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 Rollback Detector boundary before extracting or scaling it.
Range Allocator	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Stop generation on lost lease and monitor sampled/full uniqueness at system boundaries.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Rollback Detector from also owning the contract and failure behavior of ID Decoder Tool.	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 “Identifiers leak business volume” to Rollback Detector.	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 Range Allocator boundary before extracting or scaling it.
ID Decoder Tool	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Acquire a unique worker/region identity through consensus and persist generation locally.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Range Allocator from also owning the contract and failure behavior of Collision Monitor.	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 “Clock moves backward” to Range Allocator.	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 ID Decoder Tool boundary before extracting or scaling it.
Collision Monitor	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Read monotonic time; detect wall-clock rollback before composing an ID.	Proves correctness and supports controlled recovery. It prevents ID Decoder Tool from also owning the contract and failure behavior of Metrics / Capacity.	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 “Two processes share worker ID” to ID Decoder Tool.	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 Collision Monitor boundary before extracting or scaling it.
Metrics / Capacity	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Pack timestamp, region, worker, and per-millisecond sequence into configured bits.	Proves correctness and supports controlled recovery. It prevents Collision Monitor from also owning the contract and failure behavior of Audit Log.	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 “Sequence overflows one millisecond” to Collision Monitor.	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 Metrics / Capacity boundary before extracting or scaling it.
Audit Log	Preserves durable ordering, replay, and consumer checkpoints. In this design it participates in: When sequence exhausts, wait for next tick or consume a leased numeric range.	Proves correctness and supports controlled recovery. It prevents Metrics / Capacity from also owning the contract and failure behavior of operators and repair tooling.	Call every downstream consumer synchronously. That couples user latency and availability to analytics, notification, indexing, and audit dependencies. Here that simpler choice would couple “Epoch/bit layout changes during rollout” to Metrics / Capacity.	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 Audit Log boundary before extracting or scaling it.

Decision 01

Chosen boundary vs simpler monolith

Snowflake-style IDs combine timestamp, worker ID, and sequence for compact sortable uniqueness.

A single deployable with one transactional store is the right starting point for a small unique id generator workload. The proposed boundary becomes justified when snowflake-style ids combine timestamp, worker id, and sequence for compact sortable uniqueness. Separating it contains the named failure “Clock moves backward” without placing unrelated work on the critical path.

Decision 02

Correctness vs availability

Lease worker IDs through consensus and stop generation if the lease is lost; duplicate workers are catastrophic.

The design coordinates only where each accepted unique id generator 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

Handle clock rollback by waiting, using a logical epoch, or switching to monotonic sequence ranges.

This choice adds dashboards, ownership, schema compatibility, capacity planning, and recovery runbooks. Start with the minimum topology that supports “Pack timestamp, region, worker, and per-millisecond sequence into configured bits.” 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 unique id generator 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

Snowflake-style IDs combine timestamp, worker ID, and sequence for compact sortable uniqueness.

Scale boundary

Lease worker IDs through consensus and stop generation if the lease is lost; duplicate workers are catastrophic.

Operational boundary

Handle clock rollback by waiting, using a logical epoch, or switching to monotonic sequence ranges.

↑ top

5. Critical flow and algorithm

Acquire a unique worker/region identity through consensus and persist generation locally.

Read monotonic time; detect wall-clock rollback before composing an ID.

Pack timestamp, region, worker, and per-millisecond sequence into configured bits.

When sequence exhausts, wait for next tick or consume a leased numeric range.

Stop generation on lost lease and monitor sampled/full uniqueness at system boundaries.

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

Cache immutable or slowly changing reads with versioned keys, TTL jitter, short negative entries, and request coalescing.

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 unique id generator 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
Clock moves backward	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.
Two processes share worker ID	Resolve at source of truth with uniqueness, expected version, lease generation, or monotonic cursor.	Reconciliation returns zero invariant gaps; lag/replica health and user-visible SLO return to baseline.
Sequence overflows one millisecond	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.
Epoch/bit layout changes during rollout	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.
Identifiers leak business volume	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: Clock moves backward
operation_id: op_01J...
aggregate_key: unique-id-generator_123
last_durable_version: 18
projection_checkpoint: partition-07@offset-482901
retry_policy: same_identity / bounded_backoff
reconciliation_status: pending
owner: Audit Log

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: Clock moves backward — 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

UUID or Snowflake—which and why?

Snowflake-style IDs combine timestamp, worker ID, and sequence for compact sortable uniqueness. Walk the concrete lifecycle step “Acquire a unique worker/region identity through consensus and persist generation locally.” Then test it against “Clock moves backward”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

What happens on clock rollback?

Lease worker IDs through consensus and stop generation if the lease is lost; duplicate workers are catastrophic. Walk the concrete lifecycle step “Read monotonic time; detect wall-clock rollback before composing an ID.” Then test it against “Two processes share worker ID”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How are worker IDs assigned?

Handle clock rollback by waiting, using a logical epoch, or switching to monotonic sequence ranges. Walk the concrete lifecycle step “Pack timestamp, region, worker, and per-millisecond sequence into configured bits.” Then test it against “Sequence overflows one millisecond”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

Can IDs reveal business volume?

Snowflake-style IDs combine timestamp, worker ID, and sequence for compact sortable uniqueness. Walk the concrete lifecycle step “When sequence exhausts, wait for next tick or consume a leased numeric range.” Then test it against “Epoch/bit layout changes during rollout”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How do regions remain unique?

Lease worker IDs through consensus and stop generation if the lease is lost; duplicate workers are catastrophic. Walk the concrete lifecycle step “Stop generation on lost lease and monitor sampled/full uniqueness at system boundaries.” Then test it against “Identifiers leak business volume”. 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 unique id generator 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 unique id generator 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

Service Client

Regional Endpoint

ID Worker

Monotonic Clock

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