1 min lesson
Multi-region and multi-cloud geo-deployment
Describe what "Going multi-region is a tax you pay for two things" changes in practice.
Step 1 of 3
Going multi-region is a tax you pay for two things: lower latency for distant users and survival when a region fails. Pick the topology that buys the resilience you actually need, not the most impressive diagram.
The first fork is active-active versus active-passive. The honest framing is cost and operational complexity against how fast you must recover. Lay it out before you commit.
Learn more
Advanced table
Active-active for latency and instant failover
- Traffic
- Active-active
- All regions serve live traffic
- Active-passive
- One region serves; others stand by
- Failover time
- Active-active
- Near-zero; just stop steering to the dead region
- Active-passive
- Minutes; promote standby, warm caches, repoint DNS
- Cost
- Active-active
- High - full capacity running everywhere
- Active-passive
- Lower - standby can be smaller or scaled down
- Data complexity
- Active-active
- Hard - multi-writer conflicts, replication lag matters
- Active-passive
- Simpler - single writer, async replica promoted on failover
- Best when
- Active-active
- Latency for a global user base is first-order
- Active-passive
- Resilience matters but a few minutes of failover is tolerable
| Active-active | Active-passive | |
|---|---|---|
| Traffic | All regions serve live traffic | One region serves; others stand by |
| Failover time | Near-zero; just stop steering to the dead region | Minutes; promote standby, warm caches, repoint DNS |
| Cost | High - full capacity running everywhere | Lower - standby can be smaller or scaled down |
| Data complexity | Hard - multi-writer conflicts, replication lag matters | Simpler - single writer, async replica promoted on failover |
| Best when | Latency for a global user base is first-order | Resilience matters but a few minutes of failover is tolerable |
Active-active for latency and instant failover; active-passive when a controlled few-minute recovery is acceptable for far less cost and complexity.