Examkin

Question

Difficulty: Very hardManaging Cloud Run Resources

To accommodate an upcoming peak workload, a cloud infrastructure engineer must release a performance-tuned container image `gcr.io/prod-project/event-handler:v2` as a new revision for an existing Cloud Run service named `event-handler`. The operational specification demands that the new container revision must maintain a minimum of 5 warm instances, process up to 250 concurrent requests per container instance, and initially receive exactly 15% of live production traffic, with the remaining 85% remaining assigned to the baseline revision `event-handler-v1`. Which command execution workflow correctly implements this deployment without inadvertently routing 100% of live traffic to the new revision upon deployment?

  1. A
    Execute `gcloud run deploy event-handler --image gcr.io/prod-project/event-handler:v2 --concurrency 250 --min-instances 5` followed by `gcloud run services update-traffic event-handler --to-revisions event-handler-v1=85,LATEST=15`
  2. B
    Execute `gcloud run deploy event-handler --image gcr.io/prod-project/event-handler:v2 --port 250 --min-instances 5 --no-traffic` followed by `gcloud run services update-traffic event-handler --to-revisions event-handler-v1=85,event-handler-v2=15`
  3. Execute `gcloud run deploy event-handler --image gcr.io/prod-project/event-handler:v2 --concurrency 250 --min-instances 5 --no-traffic` followed by `gcloud run services update-traffic event-handler --to-revisions event-handler-v1=85,event-handler-v2=15`Answer
  4. D
    Execute `gcloud functions deploy event-handler --image gcr.io/prod-project/event-handler:v2 --concurrency 250 --min-instances 5` followed by `gcloud run services update-traffic event-handler --to-revisions event-handler-v1=85,event-handler-v2=15`

Answer

Deploy the new revision using `gcloud run deploy event-handler --image gcr.io/prod-project/event-handler:v2 --concurrency 250 --min-instances 5 --no-traffic`, then update the traffic allocation using `gcloud run services update-traffic event-handler --to-revisions event-handler-v1=85,event-handler-v2=15`.
The correct option deploys the revision while suppressing immediate traffic assignment via `--no-traffic`, configures container concurrency (`--concurrency 250`) and minimum warm instances (`--min-instances 5`), and subsequently executes `gcloud run services update-traffic` to split live traffic between `event-handler-v1` (85%) and `event-handler-v2` (15%).

Step-by-Step Solution

1
Deploy the new container image revision with configuration flags and traffic prevention.
Revision `event-handler-v2` is created with 250 max concurrency and 5 minimum warm instances, receiving 0% traffic.
The `--no-traffic` flag prevents Cloud Run's default behavior of automatically directing 100% of live traffic to newly deployed revisions.
2
Execute the traffic update command referencing the explicit revision identifiers and percentage split.
Production traffic is updated so that `event-handler-v1` serves 85% and `event-handler-v2` serves 15%.
The `gcloud run services update-traffic` command allows precise canary traffic splitting across revision names.

Key Concept

Cloud Run Revision Management, Concurrency Settings, and Traffic Splitting
Rate this question

Topics

Managing Cloud Run Resources

More questions from this topic