05Design patternsDistributed caching
05 · Reusable pattern
Distributed caching
Combine placement, replication, invalidation, eviction, and stampede protection.Lesson spine
What you need to understand.
A distributed cache adds placement, membership, replication, and origin-protection problems to ordinary caching.
Placement
Use client-side or proxy routing with consistent hashing so membership changes move a bounded fraction of keys.
Membership epochs
Version the ring and let clients handle transitions without sending one key to two conflicting owners indefinitely.
Replication
Keep copies across failure domains when cache availability matters; decide whether replicas may serve stale versions.
Eviction
Choose a policy by workload and value size, reserve headroom, and expose memory pressure before churn collapses hit rate.
Hot keys and stampedes
Replicate popular values, use local L1 caches, grant one fill lease, and serve stale while refreshing when safe.
Rebalancing
Move ranges gradually, shadow-read during cutover, and cap refill bandwidth to protect the origin.
Before the boxes
Frame the decision.
What must work
Combine placement, replication, invalidation, eviction, and stampede protection.
What changes the design
Key count · value size · popularity skew · replicas · churn · origin capacity
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.
Computes placement
Publishes ownership
Stores objects
Covers node loss
Removes stale versions
Collapses misses
Moves ranges
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
Client — Computes placement Define the output contract before moving to the next owner.
- 02
Membership — Publishes ownership Define the output contract before moving to the next owner.
- 03
Cache shard — Stores objects Define the output contract before moving to the next owner.
- 04
Replica — Covers node loss Define the output contract before moving to the next owner.
- 05
Invalidator — Removes stale versions Define the output contract before moving to the next owner.
- 06
Origin shield — Collapses misses Define the output contract before moving to the next owner.
- 07
Rebalancer — Moves ranges 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 | Replication improves availability but consumes memory and complicates invalidation. | 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
Membership churn can remap too many keys and flood the origin.
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 Distributed caching, the decision I want to make explicit is this: Replication improves availability but consumes memory and complicates invalidation. 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 Distributed caching, 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?