Python application for retrieving application dependency version information from containers running in Kubernetes.

This application uses Python 3.12.

1. Checkout the repository:

$ git clone [email protected]:dsteelma-umd/application-version-inventory.git
$ cd application-version-inventory

2. Build the Docker image:

$  docker build -t app-version-inventory:latest -f Dockerfile .

3. Switch to the Kubernetes "test" namespace:

$ kubectl config use-context test

3. Run application using the "sample_manifest.yaml" file:

$ docker run --rm -v ~/.kube:/home/app/.kube app-version-inventory:latest \
    bash -c 'app-version-inventory --namespace=test --manifest=sample_manifest.yaml'

Note: The "~/.kube" volume mount is used to provide the Kubernetes credentials to the container.

$ git clone [email protected]:dsteelma-umd/application-version-inventory.git
$ cd application-version-inventory
$ python -m venv --prompt "app-version-inventory-py$(cat .python-version)" .venv
$ source .venv/bin/activate
$ pip install -e '.[dev,test]'

Run app-version-inventory --help for the list of supported options:

$ app-version-inventory --help
Usage: app-version-inventory [OPTIONS]

Options:
  --namespace TEXT             The Kubernetes namespace to query  [required]
  --manifest FILE              The application version inventory manifest file
                               [required]
  --kube-config FILE           The ".kube/config" configuration directory
  --output-format [json|text]  The ".kube/config" configuration directory
  --help                       Show this message and exit.

The Kubernetes namespace to query, typically "test", "qa", or "prod".

The inventory manifest containing the list of applications and dependencies to query. See "Inventory Manifest" section below.

The location of "config" file containing the credentials necessary to communicate with Kubernetes. Defaults to ~/.kube/config.

The format of the output. Currently the following output options are available:

• text - Human-readable text output
• json - JSON output

The inventory manifest controls what application dependencies are queried.

See "sample_manifest.yaml" for a working example.

The general form of the manifest file is:

application:
    name: <The name of application whose dependencies are being inventoried>
    dependencies: # A list of one (or more) dependencies to query
        - name: <The name of the dependency>
          selector: <The Kubernetes label selector for the pod>
          container: <The name of the container. Optional if there is only one container>
          parser: <(Optional) The parser to use to parse the version from the "exec" command>
          exec:
            - <List of commands to run to query the version string>

The manifest may contain more than one "application" stanzas, separated by ---.

Parsers are in the "parsers" module. The following parsers are available:

• apache - Parses Apache version
• identity - Returns the "exec" command output unchanged. This is the default parser.
• java - Parses Java version
• mysql - Parses MySQL version
• node - Parses Node version
• postgres - Parses Postgres version
• python - Parses Python version
• rails - Parses Ruby on Rails version
• redis - Parses Redis version
• shibboleth_sp - Parses Shibboleth SP version
• solr - Parses Solr version
• tomcat - Parses Apache Tomcat version

New parsers can simply be added to the "parsers" module, and only require the following method:

def parse(command_response: str) -> str:
    # Do whatever processing is necessary
    return str

Formatters are in the "formatters" module. The following formatters are available:

• json - JSON formatted output
• text - returns human-readable text

New formatters can be added to the "formatters" module, and only require the following method:

def format(applications: list[Application]) -> str:
    # Do whatever processing is necessary
    return str

Also, the name of the formatter should be added to the Click "output-format" argument in "src/app_version_inventory/cli.py".