Skip to content

README.md

Latest commit

 

History

History
109 lines (69 loc) · 2.16 KB

README.md

File metadata and controls

109 lines (69 loc) · 2.16 KB

Dynamic analytics tool for AWS CLI v2

This tool traces every lines where the program actually runs during execution of AWS CLI v2.

Prerequisites

You need to install supported python versions. Details:

https://github.com/aws/aws-cli/tree/v2

For example, you can install python3.12 using pyenv for AWS CLI v2

sudo dnf install -y git gcc g++ zlib-devel bzip2 bzip2-devel readline-devel sqlite sqlite-devel openssl-devel tk-devel libffi-devel

Set up

  1. Clone repository recursively
git clone --recursive https://github.com/AWSSampleOrg/aws-cli.git
  1. Change directory to aws-cli, and checkout to the specified AWS CLI version.
cd aws-cli/aws-cli
git checkout <specific aws cli version>
  1. Change directory to the project directory, and execute following command.
cd ..
./setup.sh

How to use

  1. generate output.log

change directory to aws-cli directory.

cd aws-cli
./aws.sh s3 ls --debug
  1. Analyze with your own analyze scripts or sample scripts under analyze directory.
  • Generate html files so that those files can be seen on your web browser.
./analyze/html output.log
  • To get call stacks when the moment of specified index was executed. For example, below command shows the call stacks when index of 12000 was executed.
./analyze/call_stack output.log 12000
  • Search specified variables. For example, 169.254.169.254 were given to see IMDS related requests.
./analyze/search.sh 169.254.169.254
  • show the file content
./analyze/with_source output.log ./analyze/result.txt
  • Analyze using Athena

  1. Create database and table. Refer to analyze/query.sql. Replace <S3_BUCKET_NAME> with your own bucket.

  2. Upload the output.log to the under aws_cli_logs in the S3 bucket.

aws s3 cp output.log s3://<S3_BUCKET_NAME>/aws_cli_logs/output.log

Tips: grep

Search logging location

grep -rn 'logger.' output --include=*.py --exclude-dir=__pycache__

Clean up

  1. Change directory to the project directory.

Execute following command in the root directory.

./cleanup.sh