Which approach will allow the Specialist to review the latency, memory utilization, and CPU utilization during the load test?
Review SageMaker logs that have been written to Amazon S3 by leveraging Amazon Athena and Amazon QuickSight to visualize logs as they are being produced.
Generate an Amazon CloudWatch dashboard to create a single view for the latency, memory utilization, and CPU utilization metrics that are outputted by Amazon SageMaker.
Build custom Amazon CloudWatch Logs and then leverage Amazon ES and Kibana to query and visualize the log data as it is generated by Amazon SageMaker.
Send Amazon CloudWatch Logs that were generated by Amazon SageMaker to Amazon ES and use Kibana to query and visualize the log data.
Explanations:
While reviewing SageMaker logs in S3 with Amazon Athena and visualizing them in Amazon QuickSight is possible, this approach is not directly suited for real-time monitoring during a load test. The logs may not provide the immediate metrics needed for evaluating performance like latency, memory, and CPU utilization during the testing process.
Generating an Amazon CloudWatch dashboard is the most effective way to monitor latency, memory utilization, and CPU utilization metrics in real-time during load testing. CloudWatch provides built-in metrics from SageMaker endpoints, making it easy to visualize these performance indicators in a single view.
Building custom CloudWatch Logs and leveraging Amazon Elasticsearch Service (ES) and Kibana can provide insights into log data, but this approach is more complex and may not provide real-time performance metrics. It is not the best method for monitoring the specific metrics of latency, memory, and CPU utilization during load testing.
Sending SageMaker-generated CloudWatch Logs to Amazon ES and using Kibana to visualize the log data is not optimal for real-time monitoring of specific performance metrics such as latency, memory, and CPU utilization during a load test. This method focuses more on log analysis rather than direct performance monitoring.
From what I’ve heard, the answer is:
Generate an Amazon CloudWatch dashboard to create a single view for the latency, memory utilization, and CPU utilization metrics that are outputted by Amazon SageMaker.