HLD-316·Storage & Media
Hard · 1,500 ptsS3-Scale Multipart Blob Metadata & Placement Index
Design AWS S3 / GCS’s strongly consistent bucket-key metadata index, multipart upload commit coordinator, and erasure-coded extent locator.
Max p99
≤ 58 ms
Availability
≥ 99.5%
Budget
≤ $8,600
Functional Requirements
- FR-1PUT /buckets/{bucket_key}/complete — Atomically commits multipart extent manifest with strong read-after-write.
- FR-2GET /buckets/{bucket_key} — Resolves extent storage nodes and streams object bytes.
- FR-3PATCH & DELETE /buckets/{bucket_key}/complete — Updates, cancels, or deletes bucket_key records and immediately invalidates or updates cached state (read-after-write).
- FR-4Must provide strong read-after-write object listing/HEAD consistency and survive metadata shard crashes.
Workload Profile
Steady Cloud Workloads52,000 GET/HEAD RPS / 8,500 PUT Commit RPS
Spark/ML Checkpoint Surge138,000 Read RPS / 24,000 PUT Write RPS
Active Working Set36 GB hot bucket keyspace prefixes & extent maps
Consistency GateStrong read-after-write on PUT/DELETE (modern S3 semantics)
Consistency Model
Strong read-after-write consistency on object PUT/DELETE metadata across partitioned key ranges.
Solutions Locked
Sign in and either achieve AC or complete 3 official submissions on this question to inspect community diagrams and writeups.
Loading solutions...
Submission History (0)Select to replay
No submissions yet
Click Submit in the top bar to evaluate against all 7 judge cases and save each run here.
Design Write-up
0 boxes · 0 edges
0 words