Job Scheduler — 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 job scheduler 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 scheduled jobs · 1M due/min · seconds-level precision.
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 job scheduler 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 scheduled jobs · 1M due/min · seconds-level precision. 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 scheduled jobs · 1M due/min · seconds-level precision	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/jobscreate delayed/recurring job

PATCH /v1/jobs/{id}pause or reschedule version

POST /v1/executions/{id}/heartbeatrenew lease

GET /v1/jobs/{id}/historyexecution attempts

Authoritative records and indexes

job(job_id PK, schedule, payload_ref, priority, state, version)

execution(execution_id, job_id, scheduled_for, attempt, lease_until, state)

dedupe(job_id, scheduled_for UNIQUE, execution_id)

INDEX jobs(shard, next_run_at)

Command and event envelopes

# Command envelope
operation: POST /v1/jobs
request_id: req_01J...
idempotency_key: idem_7f2...
aggregate_key: job-scheduler_123
expected_version: 17
payload: {...}

# Durable event
event_id: evt_01J...
aggregate_key: job-scheduler_123
event_type: job_scheduler_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
Job Producer	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Persist the definition and expand only the next bounded recurrence horizon.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents the external caller from also owning the contract and failure behavior of Scheduler 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 “Scheduler leader dies after enqueue” 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 Job Producer boundary before extracting or scaling it.
Scheduler API	Owns authentication, validation, quotas, deadlines, and routing. In this design it participates in: Shard by job ID and load near-future due times into a timing wheel.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Job Producer from also owning the contract and failure behavior of Job Definition DB.	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 “Clock jump advances or delays buckets” to Job Producer.	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 Scheduler API boundary before extracting or scaling it.
Job Definition DB	Owns authoritative versioned state or immutable evidence. In this design it participates in: At due time create one execution identity, claim a lease, and enqueue by priority.	Defines the source of truth and commit boundary. It prevents Scheduler API from also owning the contract and failure behavior of Timing Wheel.	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 “Long job loses lease during GC pause” to Scheduler API.	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 Job Definition DB boundary before extracting or scaling it.
Timing Wheel	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Workers heartbeat, execute idempotently, then conditionally mark success or retry.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Job Definition DB from also owning the contract and failure behavior of Due-Time Index.	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 “Cron expansion creates duplicate occurrence” to Job Definition DB.	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 Timing Wheel boundary before extracting or scaling it.
Due-Time Index	Serves a low-latency, disposable projection with authoritative fallback. In this design it participates in: Expired leases requeue the same execution identity; DLQ preserves terminal failures.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Timing Wheel from also owning the contract and failure behavior of Shard Leader.	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 “Priority tenant starves background work” to Timing Wheel.	Costs invalidation, staleness, stampede protection, and memory. Add or split it only after the workload (1B scheduled jobs · 1M due/min · seconds-level precision) proves repeated reads dominate. Measure the latency, saturation, lag, or mismatch at the Due-Time Index boundary before extracting or scaling it.
Shard Leader	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Persist the definition and expand only the next bounded recurrence horizon.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Due-Time Index from also owning the contract and failure behavior of Lease / Claim 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 “Scheduler leader dies after enqueue” to Due-Time Index.	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 Shard Leader boundary before extracting or scaling it.
Lease / Claim Store	Owns authoritative versioned state or immutable evidence. In this design it participates in: Shard by job ID and load near-future due times into a timing wheel.	Defines the source of truth and commit boundary. It prevents Shard Leader from also owning the contract and failure behavior of Dispatch Queue.	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 “Clock jump advances or delays buckets” to Shard Leader.	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 Lease / Claim Store boundary before extracting or scaling it.
Dispatch Queue	Preserves durable ordering, replay, and consumer checkpoints. In this design it participates in: At due time create one execution identity, claim a lease, and enqueue by priority.	Decouples bursts and failure from user acknowledgment. It prevents Lease / Claim Store from also owning the contract and failure behavior of Priority Router.	Call every downstream consumer synchronously. That couples user latency and availability to analytics, notification, indexing, and audit dependencies. Here that simpler choice would couple “Long job loses lease during GC pause” to Lease / Claim Store.	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 Dispatch Queue boundary before extracting or scaling it.
Priority Router	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Workers heartbeat, execute idempotently, then conditionally mark success or retry.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Dispatch Queue from also owning the contract and failure behavior of Worker Pool.	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 “Cron expansion creates duplicate occurrence” to Dispatch Queue.	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 Priority Router boundary before extracting or scaling it.
Worker Pool	Executes idempotent asynchronous work under lease and retry policy. In this design it participates in: Expired leases requeue the same execution identity; DLQ preserves terminal failures.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Priority Router from also owning the contract and failure behavior of Execution Heartbeats.	Perform the work before acknowledging the caller. This gives simpler consistency but unbounded tail latency and fragile retries. Here that simpler choice would couple “Priority tenant starves background work” to Priority Router.	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 Pool boundary before extracting or scaling it.
Execution Heartbeats	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Persist the definition and expand only the next bounded recurrence horizon.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Worker Pool from also owning the contract and failure behavior of Retry / DLQ.	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 “Scheduler leader dies after enqueue” to Worker Pool.	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 Execution Heartbeats boundary before extracting or scaling it.
Retry / DLQ	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Shard by job ID and load near-future due times into a timing wheel.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Execution Heartbeats from also owning the contract and failure behavior of Cron Expander.	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 jump advances or delays buckets” to Execution Heartbeats.	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 Retry / DLQ boundary before extracting or scaling it.
Cron Expander	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: At due time create one execution identity, claim a lease, and enqueue by priority.	Separates a distinct scale, ownership, policy, or failure boundary. It prevents Retry / DLQ from also owning the contract and failure behavior of History / Audit.	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 “Long job loses lease during GC pause” to Retry / DLQ.	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 Cron Expander boundary before extracting or scaling it.
History / Audit	Owns one bounded domain decision, transition, or operational capability. In this design it participates in: Workers heartbeat, execute idempotently, then conditionally mark success or retry.	Proves correctness and supports controlled recovery. It prevents Cron Expander 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 “Cron expansion creates duplicate occurrence” to Cron Expander.	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 History / Audit boundary before extracting or scaling it.

Decision 01

Chosen boundary vs simpler monolith

Bucket jobs by due time, scan a short horizon, and claim work using leases or conditional updates.

A single deployable with one transactional store is the right starting point for a small job scheduler workload. The proposed boundary becomes justified when bucket jobs by due time, scan a short horizon, and claim work using leases or conditional updates. Separating it contains the named failure “Scheduler leader dies after enqueue” without placing unrelated work on the critical path.

Decision 02

Correctness vs availability

Require idempotent handlers; at-least-once execution is practical while exactly-once business effects need dedupe.

The design coordinates only where each accepted job scheduler 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

Separate scheduling, dispatch, and execution so backpressure in workers does not corrupt future schedules.

This choice adds dashboards, ownership, schema compatibility, capacity planning, and recovery runbooks. Start with the minimum topology that supports “At due time create one execution identity, claim a lease, and enqueue by priority.” 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 job scheduler 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

Bucket jobs by due time, scan a short horizon, and claim work using leases or conditional updates.

Scale boundary

Require idempotent handlers; at-least-once execution is practical while exactly-once business effects need dedupe.

Operational boundary

Separate scheduling, dispatch, and execution so backpressure in workers does not corrupt future schedules.

↑ top

5. Critical flow and algorithm

Persist the definition and expand only the next bounded recurrence horizon.

Shard by job ID and load near-future due times into a timing wheel.

At due time create one execution identity, claim a lease, and enqueue by priority.

Workers heartbeat, execute idempotently, then conditionally mark success or retry.

Expired leases requeue the same execution identity; DLQ preserves terminal failures.

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 job scheduler 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
Scheduler leader dies after enqueue	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.
Clock jump advances or delays buckets	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.
Long job loses lease during GC pause	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.
Cron expansion creates duplicate occurrence	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.
Priority tenant starves background work	Isolate the key/tenant/priority, apply admission control, and preserve durable work.	Reconciliation returns zero invariant gaps; lag/replica health and user-visible SLO return to baseline.

Recovery checkpoint

# Recovery checkpoint
scenario: Scheduler leader dies after enqueue
operation_id: op_01J...
aggregate_key: job-scheduler_123
last_durable_version: 18
projection_checkpoint: partition-07@offset-482901
retry_policy: same_identity / bounded_backoff
reconciliation_status: pending
owner: History / Audit

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: Scheduler leader dies after enqueue — 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 cron jobs avoid duplicate runs?

Bucket jobs by due time, scan a short horizon, and claim work using leases or conditional updates. Walk the concrete lifecycle step “Persist the definition and expand only the next bounded recurrence horizon.” Then test it against “Scheduler leader dies after enqueue”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How do you recover expired leases?

Require idempotent handlers; at-least-once execution is practical while exactly-once business effects need dedupe. Walk the concrete lifecycle step “Shard by job ID and load near-future due times into a timing wheel.” Then test it against “Clock jump advances or delays buckets”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How do you support priorities?

Separate scheduling, dispatch, and execution so backpressure in workers does not corrupt future schedules. Walk the concrete lifecycle step “At due time create one execution identity, claim a lease, and enqueue by priority.” Then test it against “Long job loses lease during GC pause”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

What if clocks disagree?

Bucket jobs by due time, scan a short horizon, and claim work using leases or conditional updates. Walk the concrete lifecycle step “Workers heartbeat, execute idempotently, then conditionally mark success or retry.” Then test it against “Cron expansion creates duplicate occurrence”. State the retry identity, durable checkpoint, recovery signal, and cost of the alternative.

How do you pause a tenant safely?

Require idempotent handlers; at-least-once execution is practical while exactly-once business effects need dedupe. Walk the concrete lifecycle step “Expired leases requeue the same execution identity; DLQ preserves terminal failures.” Then test it against “Priority tenant starves background work”. 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 job scheduler 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 job scheduler 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

Job Producer

Scheduler API

Job Definition DB

Timing Wheel

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