Configuration Management — 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 configuration management 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 Millions clients · updates in seconds · strong audit requirements.
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 configuration management 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 Millions clients · updates in seconds · strong audit requirements. Convert it into peak throughput, retained records, bandwidth, and failure budget; then connect each number to a component.

Dimension	Working estimate	Design implication
Traffic	Millions clients · updates in seconds · strong audit requirements	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/configs/{name}/versionsimmutable candidate

POST /v1/environments/{env}/promoteatomic pointer

GET /v1/snapshots/{version}checksum bundle

WATCH /v1/environments/{env}?after=ordered updates

Authoritative records and indexes

config_version(name, version, schema_version, payload, checksum)

environment_pointer(env, name, active_version, generation)

rollout(id, target_version, cohorts, current_stage, state)

client_ack(client_id, env, version, observed_at)

Command and event envelopes

# Command envelope
operation: POST /v1/configs/{name}/versions
request_id: req_01J...
idempotency_key: idem_7f2...
aggregate_key: configuration-service_123
expected_version: 17
payload: {...}

# Durable event
event_id: evt_01J...
aggregate_key: configuration-service_123
event_type: configuration_service_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
Operator / CI	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Authenticate/authorize author, validate schema, references, and organization policy.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents the external caller from also owning the contract and failure behavior of Config API.	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 “Schema-valid config is operationally unsafe” 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 Operator / CI boundary before extracting or scaling it.
Config API	Owns authentication, validation, quotas, deadlines, and routing. In this design it participates in: Store immutable candidate and produce a semantic diff for approval.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Operator / CI from also owning the contract and failure behavior of RBAC / Approval.	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 “Relay misses ordered update” to Operator / CI.	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 Config API boundary before extracting or scaling it.
RBAC / Approval	Owns client protocol, request identity, retry, and resume behavior. In this design it participates in: Canary by cohort; clients fetch checksum bundle and atomically swap local state.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Config API from also owning the contract and failure behavior of Schema Validator.	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 “Client cannot parse new schema” to Config API.	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 RBAC / Approval boundary before extracting or scaling it.
Schema Validator	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Observe health/acknowledgment gates, then advance or roll back environment pointer.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents RBAC / Approval from also owning the contract and failure behavior of Policy 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 “Global pointer advances too quickly” to RBAC / Approval.	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 Schema Validator boundary before extracting or scaling it.
Policy Engine	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Disconnected clients resume watch from generation or poll the complete snapshot.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Schema Validator from also owning the contract and failure behavior of Immutable Version Store.	Embed policy in every caller. Rules then drift by deployment version and emergency changes require fleet-wide redeployment. Here that simpler choice would couple “Secret reference resolves in wrong tenant” to Schema Validator.	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 Policy Engine boundary before extracting or scaling it.
Immutable Version Store	Owns authoritative versioned state or immutable evidence. In this design it participates in: Authenticate/authorize author, validate schema, references, and organization policy.	Defines the source of truth and commit boundary. It prevents Policy Engine from also owning the contract and failure behavior of Environment Pointer.	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 “Schema-valid config is operationally unsafe” to Policy Engine.	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 Immutable Version Store boundary before extracting or scaling it.
Environment Pointer	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Store immutable candidate and produce a semantic diff for approval.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Immutable Version Store from also owning the contract and failure behavior of Rollout Controller.	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 “Relay misses ordered update” to Immutable Version 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 Environment Pointer boundary before extracting or scaling it.
Rollout Controller	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Canary by cohort; clients fetch checksum bundle and atomically swap local state.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Environment Pointer from also owning the contract and failure behavior of Change Log.	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 “Client cannot parse new schema” to Environment Pointer.	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 Rollout Controller boundary before extracting or scaling it.
Change Log	Preserves durable ordering, replay, and consumer checkpoints. In this design it participates in: Observe health/acknowledgment gates, then advance or roll back environment pointer.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Rollout Controller from also owning the contract and failure behavior of Relay Root.	Call every downstream consumer synchronously. That couples user latency and availability to analytics, notification, indexing, and audit dependencies. Here that simpler choice would couple “Global pointer advances too quickly” to Rollout Controller.	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 Change Log boundary before extracting or scaling it.
Relay Root	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Disconnected clients resume watch from generation or poll the complete snapshot.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Change Log from also owning the contract and failure behavior of Regional Relays.	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 “Secret reference resolves in wrong tenant” to Change Log.	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 Relay Root boundary before extracting or scaling it.
Regional Relays	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Authenticate/authorize author, validate schema, references, and organization policy.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Relay Root from also owning the contract and failure behavior of Client Watch.	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 “Schema-valid config is operationally unsafe” to Relay Root.	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 Relays boundary before extracting or scaling it.
Client Watch	Owns client protocol, request identity, retry, and resume behavior. In this design it participates in: Store immutable candidate and produce a semantic diff for approval.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Regional Relays from also owning the contract and failure behavior of Local Last-Good 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 “Relay misses ordered update” to Regional Relays.	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 Client Watch boundary before extracting or scaling it.
Local Last-Good Cache	Serves a low-latency, disposable projection with authoritative fallback. In this design it participates in: Canary by cohort; clients fetch checksum bundle and atomically swap local state.	Removes repeated work from the latency-critical path. It prevents Client Watch from also owning the contract and failure behavior of Secret Reference 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 “Client cannot parse new schema” to Client Watch.	Costs invalidation, staleness, stampede protection, and memory. Add or split it only after the workload (Millions clients · updates in seconds · strong audit requirements) proves repeated reads dominate. Measure the latency, saturation, lag, or mismatch at the Local Last-Good Cache boundary before extracting or scaling it.
Secret Reference Service	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Observe health/acknowledgment gates, then advance or roll back environment pointer.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Local Last-Good Cache from also owning the contract and failure behavior of Health Gates.	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 “Global pointer advances too quickly” to Local Last-Good 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 Secret Reference Service boundary before extracting or scaling it.
Health Gates	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Disconnected clients resume watch from generation or poll the complete snapshot.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Secret Reference Service from also owning the contract and failure behavior of operators and repair tooling.	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 “Secret reference resolves in wrong tenant” to Secret Reference 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 Health Gates boundary before extracting or scaling it.

Decision 01

Chosen boundary vs simpler monolith

Store immutable versions and move environment pointers atomically after validation.

A single deployable with one transactional store is the right starting point for a small configuration management workload. The proposed boundary becomes justified when store immutable versions and move environment pointers atomically after validation. Separating it contains the named failure “Schema-valid config is operationally unsafe” without placing unrelated work on the critical path.

Decision 02

Correctness vs availability

Distribute changes through watches but support polling and last-known-good startup paths.

The design coordinates only where each accepted configuration management 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

Roll out progressively by cohort; automatically halt or roll back when health signals regress.

This choice adds dashboards, ownership, schema compatibility, capacity planning, and recovery runbooks. Start with the minimum topology that supports “Canary by cohort; clients fetch checksum bundle and atomically swap local state.” 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 configuration management 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

Store immutable versions and move environment pointers atomically after validation.

Scale boundary

Distribute changes through watches but support polling and last-known-good startup paths.

Operational boundary

Roll out progressively by cohort; automatically halt or roll back when health signals regress.

↑ top

5. Critical flow and algorithm

Authenticate/authorize author, validate schema, references, and organization policy.

Store immutable candidate and produce a semantic diff for approval.

Canary by cohort; clients fetch checksum bundle and atomically swap local state.

Observe health/acknowledgment gates, then advance or roll back environment pointer.

Disconnected clients resume watch from generation or poll the complete snapshot.

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 configuration management 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
Schema-valid config is operationally unsafe	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.
Relay misses ordered update	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.
Client cannot parse new schema	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.
Global pointer advances too quickly	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.
Secret reference resolves in wrong tenant	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: Schema-valid config is operationally unsafe
operation_id: op_01J...
aggregate_key: configuration-service_123
last_durable_version: 18
projection_checkpoint: partition-07@offset-482901
retry_policy: same_identity / bounded_backoff
reconciliation_status: pending
owner: Health Gates

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: Schema-valid config is operationally unsafe — 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

How do watches survive disconnects?

Store immutable versions and move environment pointers atomically after validation. Walk the concrete lifecycle step “Authenticate/authorize author, validate schema, references, and organization policy.” Then test it against “Schema-valid config is operationally unsafe”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How do you validate schemas?

Distribute changes through watches but support polling and last-known-good startup paths. Walk the concrete lifecycle step “Store immutable candidate and produce a semantic diff for approval.” Then test it against “Relay misses ordered update”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How do secrets differ from config?

Roll out progressively by cohort; automatically halt or roll back when health signals regress. Walk the concrete lifecycle step “Canary by cohort; clients fetch checksum bundle and atomically swap local state.” Then test it against “Client cannot parse new schema”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How do you prevent a bad global rollout?

Store immutable versions and move environment pointers atomically after validation. Walk the concrete lifecycle step “Observe health/acknowledgment gates, then advance or roll back environment pointer.” Then test it against “Global pointer advances too quickly”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How do clients pin a version?

Distribute changes through watches but support polling and last-known-good startup paths. Walk the concrete lifecycle step “Disconnected clients resume watch from generation or poll the complete snapshot.” Then test it against “Secret reference resolves in wrong tenant”. 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 configuration management 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 configuration management 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

Operator / CI

Config API

RBAC / Approval

Schema Validator

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