Exercises › Progressive Delivery Language › Blue-Green Deployment Vocabulary

Blue-Green Deployment Vocabulary

5 questions · Progressive Delivery Language

1. An engineer describes their release process: "We deploy the new version to the green environment while blue serves all production traffic. After testing green, we flip the router." What defines a blue-green deployment?

Two deployment pipelines running in parallel: blue deploys on Mondays, green deploys on Thursdays Two identical production environments — one serving live traffic (active), one idle (inactive) — the new version is deployed to the idle environment and traffic is switched over instantly when ready A two-stage deployment where the new version is gradually rolled out to blue team users first, then to green team users

2. A team keeps the blue environment running for 24 hours after switching traffic to green. Why is this practice recommended?

To serve as a staging environment for the next deployment cycle To preserve instant rollback capability — if a problem is discovered in green after cutover, traffic can be switched back to blue immediately without a redeployment Compliance requirements mandate that two production-equivalent environments run simultaneously at all times

3. A team plans a blue-green deployment that includes a database schema change. The architect warns: "We need to ensure schema compatibility." Why is this a specific challenge for blue-green deployments?

Blue-green deployments require separate databases for each environment, making schema migrations twice as expensive During and after cutover, both the old (blue) and new (green) application versions may need to read from and write to the same database — so any schema change must be backward-compatible with both versions simultaneously Blue-green deployments cannot support schema migrations at all — schema changes require a big-bang deployment with downtime

4. An engineer describes the traffic cutover step: "We update the load balancer to point to the green target group." What happens to in-flight requests during this switch?

All in-flight requests are queued and replayed against the green environment after the switch Requests already in flight to blue complete normally; new connections after the switch go to green — modern load balancers handle this gracefully with connection draining All active user sessions are terminated and users must re-authenticate against the green environment

5. A newly deployed green environment receives traffic but performs poorly for the first few minutes. The engineer says: "It needs a warm-up period." What causes this and how is it addressed?

The green environment needs time for DNS propagation to complete before it can fully serve traffic Newly started application instances have cold caches (JIT compilation, in-memory caches, connection pools) — performance is lower until these warm up; addressed by pre-warming or sending synthetic traffic before cutover Load balancer health checks take several minutes to confirm the green environment is healthy before routing full traffic

Exercise complete!

out of 5 questions

← Back to Progressive Delivery Language All exercises