Global low-latency at scale

expert

Scenario & brief

A worldwide audience needs p99 ≤ 85 ms at 8,000 rps. No matter how you tune a single region, distant users pay the round-trip tax — you simply can't hit 85 ms globally from one place.

Targets: p99 ≤ 85 ms, 99.95% availability, critical durability, under $2,000/mo.

Serve users close to home: a CDN for edge caching plus an active-active multi-region deployment to cut round-trip latency, backed by a right-sized autoscaling tier, a cache, and a Multi-AZ database. Multi-region doubles the compute/database bill, so keep every tier lean.

8,000 rps peakp99 ≤ 85ms99.95% availdurability: criticalbudget $2,000/mo

CloudFront

Networking

Enable CDN

System health

Erupting · SLA breach

/ 100

Score

SLA not met yet

Monthly cost

$3,755

Budget $2,000/mo · $1,755 over

Metrics

Capacity43

Availability30

Durability25

Cost efficiency7

Requirements

Peak capacity — 39000 rps compute · 3400 rps db (need ≥ 8000 rps)
p99 latency — ~134 ms (need ≤ 85 ms)
Availability — 99.00% (need ≥ 99.95%)
Durability — at risk (need redundancy + backups)
Budget — $3755/mo (need ≤ $2000/mo)

Advisor

The database is saturated at peak — add a cache to shed read load, or scale it up.
Compute runs in a single AZ — spread across ≥2 AZs (with ≥2 instances) to meet the availability target.
The database has no Multi-AZ standby or replica — a failure risks data loss. Enable Multi-AZ and backups.

Discussion

No comments yet. Be the first to start the discussion.

For learning purposes only. Costs and capacities are illustrative, not live AWS prices. Not affiliated with or endorsed by Amazon Web Services.