An CLI tool for the Opsani Optimization API: coctl

This is a simple application that can consume an Opsani service configuration and post it to the domain/app, or query the service for the current domain/app configurations.

Work in progress

Basic get (download current configuration), and put (apply a patch to the upstream config, merge with what's there) is aavailable. there's also a restart function, but that could also be handled by passing a null '{}' document to put.

Inputs are not validated Error handling could certainly be improved

Installation

It is recommended that you launch this in a Docker container, or deploy in a python virtual environment:

python3 -m venv .
. bin/activate
python3 -m pip install -e .

or:

docker build . -t coctl:latest
alias coctl='docker run -it --rm --name coctl -v $(pwd)/:/work coctl:latest '

Basic usage

Export the configuration (or pass as CLI parameters):

export CO_TOKEN=ASFDASDFASDFASDF
export CO_DOMAIN=domain.name
export CO_APP=app.name

For Docker use, set an alias after setting the environment variables, and you won't have to pass the env variables:

alias coctl="docker run -it --rm --name coctl -v $(pwd)/:/work/ -e CO_TOKEN=$CO_TOKEN -e CO_DOMAIN=$CO_DOMAIN -e CO_APP=$CO_APP coctl:latest "

Get config

Get the current configuration from your domain/app:

coctl get

Get the current config to a specific file:

coctl get -f config-state.yaml

Put config

Put a YAML config in coconfig.yaml to the app config first get the config, then re-put it to the application endpoint.

coctl get
coctl put

For a YAML document with a specific perf: metric, update the optimization: configuration

cat > optimization.yaml <<EOF
optimization:
  mode: saturation
  perf: metrics['requests throughput']
EOF

coctl put -f optimization.yaml

Change just the optimiization

Restart the optimization:

coctl restart

TODO

1. Basic configuration validation
2. Basic environment variable validation
3. Remove the need for the --file option, treat as an argument instead