如何将Jenkins输出保存至文件并通过Shell命令上传至S3（无需构建后插件）

1. How to save Jenkins output to a file? #

Jenkins has several simple ways to preserve build output:

• Built-in log storage: Jenkins automatically keeps console logs for every build (you can view them via the build's "Console Output" page). If you need a persistent copy stored on the Jenkins server, use the Archive the artifacts post-build action (no extra plugins required for basic archiving) to save log files or generated output artifacts.

• Shell command redirection: During your build steps, redirect output from individual commands or entire scripts to a file directly. Examples:

  # Redirect only standard output to a file (overwrites existing content)
  ./your-build-script.sh > build-output.log
  
  # Redirect both standard output and error to the same file
  ./your-build-script.sh &> build-output.log
  
  # Append output to a file instead of overwriting
  ./your-build-script.sh >> build-output.log
  

2. Save Jenkins output to an S3-uploadable file without post-build plugins #

If you want to avoid post-build plugins and handle everything via shell commands, here's a practical method using the AWS CLI (ensure your Jenkins agent has AWS CLI installed and configured with S3 write permissions):

Step 1: Capture build output to a file #

First, capture all output from your build commands—including errors—to a log file for complete visibility:

# Enable pipefail to ensure the script fails if any command inside fails
set -o pipefail  

{
  echo "=== Build started at $(date) ==="
  # Run your actual build commands here
  npm install
  npm run build
  echo "=== Build finished at $(date) ==="
} &> build-output.log

Step 2: Upload the file to S3 #

Use the AWS CLI s3 cp command to upload the log file to your S3 bucket. Add the Jenkins build number to the filename to avoid overwriting previous logs:

# Upload to S3 with unique build identifier
aws s3 cp build-output.log s3://your-bucket-name/jenkins-build-logs/build-${BUILD_NUMBER}-output.log

# Optional: Set access controls (e.g., make the log publicly readable)
aws s3 cp build-output.log s3://your-bucket-name/jenkins-build-logs/build-${BUILD_NUMBER}-output.log --acl public-read

Key Notes: #

• ${BUILD_NUMBER} is a Jenkins environment variable that auto-replaces with the current build's number—great for organizing logs chronologically.
• Configure AWS credentials on the Jenkins agent via environment variables, ~/.aws/credentials, or IAM roles (if running on EC2 instances) to avoid hardcoding secrets.
• If you need the full Jenkins console log (not just your shell step output), you can fetch it via the Jenkins API in a shell step, but this requires authentication. For most use cases, capturing build step output directly is sufficient.

内容的提问来源于stack exchange，提问作者Lawrence Johnson