Run the status page in both AWS and GCP and put a health-checked DNS layer in front so traffic serves from the healthy provider and fails over to the other within minutes when one goes dark — no human paging required.
#dns#resilience#networking
LabPingdread, a status-page company whose own status page is single-cloud — a fact that became deeply embarrassing the day its only provider had a regional incident and customers had nowhere to check whether anything was down.all labs
02 - Actions
score -- - -- votes
-- completed
Sign in with a real account to save ratings, completions, and share snapshots.
Stateloading
03 - Scenario
Pingdread, a status-page company whose own status page is single-cloud — a fact that became deeply embarrassing the day its only provider had a regional incident and customers had nowhere to check whether anything was down.
Run the status page in both AWS and GCP and put a health-checked DNS layer in front so traffic serves from the healthy provider and fails over to the other within minutes when one goes dark — no human paging required.
Constraints
Active in both clouds; neither is a cold standby that needs manual promotion
Automatic cutover under 5 minutes on a provider-down health check
DNS layer and both origins fully described in one Terraform root module
Failover must be demonstrable by killing one origin's health endpoint
ScenarioMULTI - advanced
05 - Steps
STEP_01
Stand up the GCP origin
Deploy the status-page origin in GCP and attach a health check that polls its health endpoint, so GCP has its own opinion about whether that origin is alive.
Define a Route 53 health check against the AWS origin's health endpoint. Tune failure_threshold and request_interval so a real outage trips inside your 5-minute budget without flapping on one slow response.
In the Route 53 zone, create a PRIMARY failover record for the AWS origin tied to its health check and a SECONDARY record pointing at the GCP origin. Set a short TTL so resolvers pick up the cutover quickly.
Hint: TTL is half the battle — a 300s record means resolvers can cache the dead primary for up to 5 minutes regardless of how fast your health check trips.
Kill the AWS origin's health endpoint, then resolve the name on a loop and watch the answer flip to the GCP origin. Record wall-clock time from "primary down" to "serving secondary" and compare it to your 5-minute RTO.