Backup Jenkins configuration to S3

Dependencies

• S3 Plugin

Global configuration

1. Set up a S3 profile at jenkinshost/configure.

Job configuration

1. Create a new free-style project and give it a name.

2. (Optional) I recommend to check “Discard Old Builds” and choose the following settings to save disk space:
   Strategy: Log Rotation
   Days to keep builds: (empty)
   Max # of builds to keep: (empty)
   Days to keep artifacts: 1
   Max # of builds to keep with artifacts: 1

3. Set a build trigger of your choice; I use this:
   Build periodically - Schedule: H H(0-5) * * *

4. Add an “Execute shell” build step with the following command:

   # Delete all files in the workspace
   rm -rf *
   # Create a directory for the job definitions
   mkdir -p $BUILD_ID/jobs
   # Copy global configuration files into the workspace
   cp $JENKINS_HOME/*.xml $BUILD_ID/
   # Copy keys and secrets into the workspace
   cp $JENKINS_HOME/identity.key $BUILD_ID/
   cp $JENKINS_HOME/secret.key $BUILD_ID/
   cp $JENKINS_HOME/secret.key.not-so-secret $BUILD_ID/
   cp -r $JENKINS_HOME/secrets $BUILD_ID/
   # Copy user configuration files into the workspace
   cp -r $JENKINS_HOME/users $BUILD_ID/
   # Copy job definitions into the workspace
   rsync -am --include='config.xml' --include='*/' --prune-empty-dirs --exclude='*' $JENKINS_HOME/jobs/ $BUILD_ID/jobs/
   # Create an archive from all copied files (since the S3 plugin cannot copy folders recursively)
   tar czf $BUILD_ID.tar.gz $BUILD_ID/
   # Remove the directory so only the archive gets copied to S3
   rm -rf $BUILD_ID

5. Create a post-build action of type “Publish artifacts to S3 Bucket” and configure as follows:
   S3 profile: Choose the profile from the global configuration
   Source: **
   Destination bucket: Enter the name of the bucket where you want the archive to go
   Note: Even though it says “Destination bucket”, it is possible to enter a bucket name AND path, the S3 plugin will create the directory or use it if it already exists.

S3 configuration

The setup described above creates a new backup each day. I like being able to go back in history, e. g. if the accidental deletion of a job was discovered after several days. I’d recommend to use a lifecycle policy on the bucket to remove old backups after a desired number of days.

Alternatively one could achieve a similar effect by enabling versioning on the bucket and change the archive’s file name to a fixed value by changing the tar command in the shell script:
tar czf jenkins-configuration.tar.gz $BUILD_ID/