Backup Teamcity Settings

Pre-requisites

Software on Ansible Machine

You will need the following installed on the machine:

• Ansible 2.4
• AWS CLI

Amazon S3

You will need to setup a S3 bucket with the appropriate settings.

Then, you should create a IAM user specifically with permissions to only put objects into the bucket. An example Policy document is this:

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Sid": "1",
            "Effect": "Allow",
            "Action": [
                "s3:ListBucket"
            ],
            "Resource": [
                "arn:aws:s3:::YOUR-BUCKET-NAME"
            ]
        },
        {
            "Sid": "2",
            "Effect": "Allow",
            "Action": [
                "s3:PutObject",
                "s3:ListMultipartUploadParts",
                "s3:AbortMultipartUpload"
            ],
            "Resource": [
                "arn:aws:s3:::YOUR-BUCKET-NAME/*"
            ]
        }
    ]
}

TeamCity

You should also setup a TeamCity user solely for the purpose of trigerring and downloading backups only. The user will need the Change backup settings and control backup process permission.

Ansible Secrets

Refer to secrets.template.yml and fill it in. By default, the playbook will look for the secrets in secrets.yml, but you can override it with the secrets_file variable.

Playbook Variables

The playbook has additional variables you can override at run time, in addition to the secrets file it will include. Refer to the playbook. You might want to refer to variable precedence to decide where to place them.

Running the entire playbook

You won't need to ask ansible to run the playbook at another host, and you can run it locally. The default invocation would simply be

ansible-playbook -i localhost --connection local site.yml

Alternatively, you can use the provided Docker Compose file to run the playbook if you have secrets.yml in the current directory:

docker-compose up

Tags

The playbook tasks occur in two separate and distinct stages that can be run independently. Each stage is tagged.

• Trigger Backup on TeamCity (trigger_backup)
• Download Backup from TeamCity and upload to S3 (upload_backup)

For instance, to only trigger backup, do

ansible-playbook -i localhost --connection local site.yml --skip-tags upload_backup

If you just want to upload a backup to S3, you will need to know the name of the backup and then do:

ansible-playbook -i localhost --connection local site.yml \
  -e "{'teamcity': { 'backup_name': 'TeamCity_Backup_2017-02-02T05-28-36Z_20170202_052905.zip'}}" \
  --skip-tags trigger_backup