Bulldog

A simple CLI Golang application which you can use to scrape your kubernetes cluster for resources you are looking for.

This tool was built with a goal to quickly take a live snapshot of the resources and their contents by querying the Kubernetes API Server and storing it in text files. The need for a tool like this arises when you have a lot of kubernetes objects which change dynamically over time and for some reasons you may want to store the changes that occur on those objects inorder to do analysis of what caused the change or how the state of an object changed over time and what caused it.

This app requires a scrapeconfig file to be provided to it and then uses the default kubeconfig file present on an instance to fetch the specified resources from the cluster which is specified in the kubeconfig file.

Using bulldog

You can either download the source code and build the application on your own machine or download the prebuilt binaries present on github releases and then execute them. Building the binary is as simple as

git clone https://github.com/gat786/bulldog
cd app && go build . -o bulldog

If you decide to download a prebuilt binary you would either need to add it to path (install) or execute it from the directory in which it is downloaded.

You can install the downloaded binary like below

# change the binary name according to your downloaded choice
mv bulldog-linux-amd64 bulldog
sudo install -m 755 bulldog /usr/local/bin

Doing this would add the binary to an executable directory and make it such that it will then be available to be executed from terminal automatically.

Executing the Bulldog CLI would then show you the supported options

$ bulldog

2026/04/27 18:26:59 info Initialising Kubernetes Resource Scraper
NAME:
   bulldog - A new cli application

USAGE:
   bulldog [global options] [command [command options]]

COMMANDS:
   scrape
   print-config
   help, h       Shows a list of commands or help for one command

GLOBAL OPTIONS:
   --help, -h  show help

There are two options as of now,

1. scrape

   When ran the Scrape option will read the default kubeconfig file present and start scraping the resources present on that cluster and saving it as a text output in the exported_data directory.

!TODO - configure a way to export to a custom user specified directory

2. print-config

   Inorder to scrape data from a cluster, this tool requires a scrapeconfig, which if not specified manually, the tool considers that you just want everything to scraped and stored. You can provide a custom scrapeconfig by specifying the CONFIG_FILE_PATH file path which when specified the yaml file on that path will be used as the scrapeconfig.

   If you are not sure what items the tool will scrape and want to see what config the tool is using you can run this command along with the environment variable specified or not and the tool will print out the config it is using to scrape the data.

Scrapeconfig

An example scrapeconfig file looks like this

namespaces:
  - default
resources:
  - Group: ""
    ResourceNames:
      - pods
      - serviceaccounts
    Version: v1
savefullmanifest: true

There are three fields which it requires to be defined in the scrapeconfig.

1. Namespaces: namespaces
2. Resources: resources
3. SaveFullManifest: savefullmanifest

As the names suggest these are namespaces from which you want the resources to be scraped, resources are the actual resources that you want to be scraped from these namespaces and the savefullmanifest flag actually asks if you want to store the yaml file currently present on api server as a backup or not. By default the scraper only lists things and will only tell you how many resources as well as what are their names in a particular cluster or namespace.