Exam Professional Cloud Developer topic 1 question 198 discussion

from documentation: "If you need to scale your workload based on the performance of an application or service outside of Kubernetes, you can configure an external metric. For example, you might need to increase the capacity of your application to ingest messages from Pub/Sub if the number of undelivered messages is trending upward." https://cloud.google.com/kubernetes-engine/docs/concepts/custom-and-external-metrics#external-metrics

Answer is C

I go with C since we need to scale GKE Deployment to be able to process messages in a TIMELY manner. External metrics is more suitable for this.

C: https://cloud.google.com/kubernetes-engine/docs/tutorials/autoscaling-metrics#pubsub

Based on the requirement that the application reads and processes Pub/Sub messages, and that the rate at which messages are published to the Pub/Sub topic varies considerably throughout the day and week, including occasional large batches of messages published at a single moment, the best choice would be C: Horizontal Pod Autoscaler based on an external metric. By using an external metric, the Horizontal Pod Autoscaler can monitor the number of messages in the Pub/Sub topic and adjust the number of replicas in the GKE Deployment accordingly. This allows the application to automatically adapt to changes in the rate at which messages are being published, ensuring that the pods are able to process messages in a timely manner. On the other hand, Horizontal Pod Autoscaler based on resources utilization, it would not provide the needed functionality as it bases scaling on resource usage of the pods, not the number of messages in the Pub/Sub topic

The answer is C for me. The need to scale is from the number of messages in the pub/sub. It's an external metrics that can be reported from pub/sub or cloud monitoring. So the scale will be better to use this metrics to add X pod (each pod is limited on the number of message per minutes so the system now how much pod you need to scale to answer at X new messages.

Custom and external metrics allow workloads to adapt to conditions besides the workload itself. Consider an application that pulls tasks from a queue and completes them. An external metric is reported from an application or service not running on your cluster, but whose performance impacts your Kubernetes application. For information, the metric could be reported from Cloud Monitoring or Pub/Sub. D isn't the answer, before selecting an answer , please do a thorough research and understand concepts and the key words in a question, D cant be the answer in this case. https://cloud.google.com/kubernetes-engine/docs/concepts/custom-and-external-metrics

The answer is C. Each pod will handle a fixed number of messages, fixed being the key word here. Now let's say this fixed number is "1000" messages per minute. Do you think a 1000 messages in a minute will cause the pod autoscaler to kick in based on resource utilisation? We need to scale using external metrics here. When Pod 1 is handling the maximum of "fixed number amount" messages, we need to spin up pod 2 etc...

option D

D is the answer. https://kubernetes.io/docs/tasks/run-application/horizontal-pod-autoscale/