Simplify CloudWatch CLI commands with Q Developer

January 29, 2025

Note: This content was originally published on the AWS Blog and is reproduced here for archival purposes. Please visit the original post for the most up-to-date version.

Welcome to another installment of Q-Bits, our regular series showcasing cool ways Amazon employees are leveraging Amazon Q Developer. Today, we're diving into how Q Developer can assist with CloudWatch in the AWS CLI.

I rely heavily on Amazon CloudWatch to monitor the health and performance of my applications. CloudWatch provides a wealth of invaluable data, but I've always found it challenging to remember the exact syntax for querying specific metrics from the command line. That is, until I started using Amazon Q Developer.

Q Developer has been a game-changer for me when it comes to interacting with CloudWatch. Instead of having to recall the various namespaces, metric names, and dimensions required for each CLI command, I can now simply describe what I want to do in natural language, and Q translates that into the appropriate AWS CLI syntax.

Let's start with an example from my work with Amazon Simple Storage Service (Amazon S3). I need to check the average size of my S3 buckets, but the full AWS CLI command to do this can be quite complex:

1
aws cloudwatch get-metric-statistics --metric-name BucketSizeBytes --namespace AWS/S3 --statistics Average --dimensions Name=BucketName,Value=my-bucket-name Name=StorageType,Value=StandardStorage --start-time 2024-12-01T00:00:00Z --end-time 2024-12-31T23:59:59Z --period 2592000 --region us-east-1 --output text --query 'Datapoints[0].Average'

With Q, I can simply say "Using CloudWatch, display the average size of the S3 bucket my-bucket-name during December 2024." Q takes care of translating that request into the correct CLI command, saving me time and headaches. Note that there a few ways to get this data at the command line, so I have specified that I want Q to use CloudWatch.

The same goes for querying other CloudWatch metrics. If I want to see the average hourly CPU utilization for a specific Amazon Elastic Compute Cloud (Amazon EC2) instance, I can say "Using CloudWatch, display the average hourly CPU utilization for instance i-0123456789abcdef on January 26, 2025." Q will handle the details of constructing the appropriate command:

1
aws cloudwatch get-metric-statistics --metric-name CPUUtilization --start-time 2025-01-26T00:00:00 --end-time 2025-01-26T23:59:59 --period 3600 --namespace AWS/EC2 --statistics Average --dimensions Name=InstanceId,Value=i-0123456789abcdef

This works for most any AWS service, like getting the average read IOPS for an Amazon Aurora cluster. I can simply say "Using CloudWatch, display the average read IOPS for the Aurora cluster named my-cluster-name on January 21, 2025," and Q will take care of the details:

1
aws cloudwatch get-metric-statistics --namespace AWS/RDS --metric-name ReadIOPS --dimensions Name=DBClusterIdentifier,Value=my-cluster-name --start-time 2025-01-21T00:00:00Z --end-time 2025-01-22T00:00:00Z --period 3600 --statistics Average

Beyond CloudWatch metrics, Q can also help with querying CloudWatch Logs. If I need to tail the logs for a Lambda function, I can just say "Using CloudWatch, tail the logs for the Lambda function my-function-name," and Q will translate that into the appropriate aws logs tail command.

After using Q for the past few years, I've found that it saves me a significant amount of time and cognitive load when working with CloudWatch. Instead of having to remember all the various service namespaces, metric names, and dimensions, I can focus on describing what I want to do in plain English, and let Q handle the technical details. This has made me much more productive and efficient in my day-to-day work with AWS services.

If you, like me, struggle to keep track of the CloudWatch CLI syntax, I highly recommend giving Amazon Q Developer a try. It's a simple but powerful tool that can help streamline your interactions with CloudWatch and other AWS services. Give it a shot, and let me know how it goes!