Basic Operations for Running UAA

This section discusses some basic methods for inspecting, debugging, scaling, and upgrading your UAA.

Whilst this section focuses on the uaa-deployment tool, most of the information is useful for any deployment of the UAA via BOSH. You can learn more about BOSH from Stark & Wayne's free online book Ultimate Guide to BOSH.

Deploy UAA to local VirtualBox

uaa-deployment up

This will create a vars.yml containing default configuration for your VirtualBox VM. You can edit it and run uaa-deployment up.

Deploy UAA to AWS

uaa-deployment up --cpi aws

This will not initially create anything on AWS, instead it will create a vars.yml file for you to edit.

access_key_id: # ...
secret_access_key: # ...
region: us-east-1
az: us-east-1b
subnet_id: # subnet-...
internal_cidr: 10.0.0.0/24
internal_gw: 10.0.0.1
internal_ip: 10.0.0.6
default_security_groups: [bosh]
default_key_name: bosh
private_key:

instance_type: m4.xlarge
ephemeral_disk_size: 25_000

# flag: --spot-instance
spot_bid_price: # Bid price in dollars for AWS spot instance

# flag: --security-groups
security_groups:

# flag: --lb-target-groups
lb_target_groups:

Once you've provided your AWS credentials, networking configuration, and even an optional spot instance price, you can then re-run the up command:

uaa-deployment up

Anytime you make changes to your vars.yml, you then apply them using uaa-deployment up. The bosh CLI will rebuild your AWS VM, install and configure the UAA software, and keep the UAA running for you on that VM.

Inside your UAA VM

Regardless if you've deployed your UAA to VirtualBox, AWS, GCP, Azure, vSphere, or another BOSH CPI infrastructure, you can SSH into your UAA VM with the same command:

uaa-deployment ssh

You can find the start scripts and configuration files for the UAA:

cd /var/vcap/jobs/uaa
cat config/uaa.yml

You can tail the logs for the various processes/sub-systems:

tail -f /var/vcap/sys/log/{*.log,*/*.log}

To see the nested set of processes running on the VM:

ps axwwf

The UAA is backed by a PostgreSQL database. Its contents are stored on a 60G persistent disk mounted at /var/vcap/store:

df -h
cd /var/vcap/store/postgres/postgres*/
ls -al

The processes are monitored using Monit. You can force it to restart all processes:

monit restart all