Disaster recovery (DR) pair diagram
A primary site and a DR site side by side, with storage replication, database log shipping and the GSLB failover on one canvas.
Open this example in the editor →
The site layer holds the primary site in Seoul and the DR site in Daejeon; the equipment layer mirrors firewalls, core switches, app and DB servers and storage on both sides. The leased line and the async storage replication (RPO 15 minutes) are written on the connections themselves, so the recovery point objective is readable straight off the picture.
The "Primary fails → switch to DR" scenario steps from normal replication to GSLB failing over and the standby database being promoted. It is a handy way to explain, in a drill plan or an audit, what moves where and in which order when the primary goes down.
What's on each layer
- L2 · Services — Users (same address), GSLB · DNS (health checks), Business app (active), Business DB (active), Business app (standby), Business DB (standby)
- L1 · Equipment — Firewall (primary), Core switch (primary), App server (primary), DB server (primary), Storage (source) (100TB), Firewall (DR), Core switch (DR), App server (DR · standby), DB server (DR · standby), Storage (replica) (100TB)
- L0 · Sites — Primary site (Seoul), DR site (Daejeon)
Flow scenario
Primary fails → switch to DR
- Storage (source) → Storage (replica) — Normally: storage replicates every 15 min
- Business DB → Business DB — DB logs shipped
- GSLB · DNS → Business app — Primary stops answering → switch to DR
- Business app → Business DB — DR app starts · standby DB promoted
- Users → GSLB · DNS — Users reconnect to the same address
Good for
- DR drill plans and post-drill reports
- Explaining redundancy for audits and certifications
- Planning a site move or a bigger link
Open this example in the editor →
More examples
- Three-tier web service diagramThe classic split — web servers in the DMZ, application and database servers inside — drawn as an equipment layer with the services stacked on top.
- 42U rack elevationOne rack with ToR switches, a firewall, a management switch, DB and app servers, storage and a UPS — each placed at its U position.
- Network separation diagramA business network and an internet network kept apart: the internet only through VDI, files only through a cross-domain transfer server.