05Design patternsJob scheduling
05 · Reusable pattern
Job scheduling
Turn future work into durable, claimable tasks with clocks, leases, fairness, and retries.Lesson spine
What you need to understand.
A scheduler converts future intent into durable, fairly claimable work despite clocks, retries, and worker failure.
Store next occurrence
Persist the schedule, timezone, misfire policy, and next fire time; do not rescan every historical expression on each tick.
Find due work
Use indexed time buckets, partitioned polling, delayed queues, or timing wheels according to timer count and precision.
Claim with a lease
A worker atomically owns an occurrence for a bounded interval and heartbeats if execution exceeds the visibility timeout.
At-least-once execution
Expect a lease to expire after ambiguous completion and require the job effect to be idempotent.
Recurring semantics
Define daylight-saving behavior, catch-up after downtime, overlap policy, and whether missed runs coalesce.
Priorities and fairness
Separate queues or use weighted admission so one tenant or long job class cannot starve the fleet.
Before the boxes
Frame the decision.
What must work
Turn future work into durable, claimable tasks with clocks, leases, fairness, and retries.
What changes the design
Timer count · precision · horizon · scan rate · runtime · missed-run policy
What owns the truth
Identify the component that commits authoritative state, then separate synchronous confirmation from derived work.
What stays simple
Do not add global coordination, multi-region writes, or a specialized store until a requirement earns the complexity.
Architecture map
Trace ownership, not just traffic.
Follow the decision from left to right. Every arrow should have a reason.
Defines schedule
Stores next occurrence
Finds due partitions
Buffers work
Claims lease
Extends execution
Records next run
Walk one representative request across every arrow. Say whether the handoff is synchronous or asynchronous, what identity makes a retry safe, and which step changes authoritative state.
- 01
User or service — Defines schedule Define the output contract before moving to the next owner.
- 02
Schedule DB — Stores next occurrence Define the output contract before moving to the next owner.
- 03
Scanner — Finds due partitions Define the output contract before moving to the next owner.
- 04
Ready queue — Buffers work Define the output contract before moving to the next owner.
- 05
Worker — Claims lease Define the output contract before moving to the next owner.
- 06
Heartbeat — Extends execution Define the output contract before moving to the next owner.
- 07
History — Records next run Confirm the result and emit the evidence needed to reconcile it.
Decision table
Make the trade-offs explicit.
| Decision | Defensible position | Cost to acknowledge |
|---|---|---|
| Primary mechanism | Indexed polling is robust; timing wheels reduce scan cost for huge timer sets. | The stronger guarantee usually adds coordination, latency, state, or operational work. |
| Sync vs. async | Keep only correctness-critical confirmation synchronous. Move derived views, notifications, analytics, and cleanup behind a durable boundary. | Async work needs idempotency, lag monitoring, replay, and a product definition for partial completion. |
| Simple vs. scaled | Begin with one logical owner and a clear API. Partition or replicate only the resource proven to be the first bottleneck. | Migration requires stable identities, versioned contracts, backfill, and a rollback path. |
Failure review
Design the recovery path.
Topic-specific risk
Clock shifts and slow scans can fire early, late, or twice.
ResponsePersist enough identity and state to distinguish retry, resume, compensation, and operator repair.
Dependency timeout
A timeout is ambiguous: the remote side may have failed, succeeded, or still be running.
ResponseUse deadlines, bounded backoff with jitter, idempotency keys, and a status or reconciliation path.
Overload or skew
Average capacity can look healthy while a tenant, key, partition, region, or expensive request saturates one owner.
ResponseExpose queue depth and hot-key share, apply backpressure, isolate tenants, and degrade optional work before correctness.
Evidence + level bar
Prove the design can be operated.
Health of the promise
Measure user-visible latency or freshness, correctness drift, saturation, retry volume, and time to recover. Alert on the failed promise—not only CPU.
Complete and clear
Finish the happy path, identify the state owner, choose reasonable building blocks, and explain one scale mechanism.
Trade-offs and failure
Separate read and write paths, define consistency, explain partitioning, and make duplicate or partial failure safe.
Evolution and operations
Discuss multi-region boundaries, migration, tenant isolation, capacity, observability, and how the architecture changes over time.
Interview language
Open the deep dive with a claim.
“For Job scheduling, the decision I want to make explicit is this: Indexed polling is robust; timing wheels reduce scan cost for huge timer sets. I’ll trace the state-changing path first, show where the result becomes durable, then test the design against the highest-risk failure and our target scale.”
08 · Retrieval check
Can you defend it without the page?
- For Job scheduling, where is the correctness boundary and which failure would you test first?
- Which component owns committed truth, and what event or response proves the commit?
- Where is the first scaling or coordination bottleneck under the stated envelope?
- What happens after an ambiguous timeout or duplicate operation?
- Which complexity would you remove at one hundredth of the scale?