Giter Site home page Giter Site logo

obsctl-reloader-rules-checker's Introduction

obsctl-reloader-rules-checker

Uploading rules to RHOBS is normally done through obsctl-reloader.

The obsctl-reloader-rules-checker tool is a command line tool checking that the PrometheusRule objects fed to obsctl-reloader can safely be consumed.

Retrieving the tool binary

You can retrieve this tool binary from GitHub release section.

The binary is delivered for all major platforms (Linux, Windows, Mac OS) and for all major architectures.

Prerequisites before using the tool binary

The obsctl-reloader-rules-checker tool relies on the following command line tools:

  • promtool: this tool is at the heart of obsctl-reloader-rules-checker.
    It is used to check that input files really store PrometheusRule objects; the tool is also used to run the rules unittests.
  • yamllint: this tool is a linter (so a nice to have) used both on the rules and the unittests.
    It only need to be present if you plan on using the -y flag of obsctl-reloader-rules-checker.
  • pint: this tool is an other linter which is aimed a tackling PrometheusRule objects.
    Again installing it is optional as its use is conditioned by the -p flag.

You have to make sure that those tool are present on your computer when using the obsctl-reloader-rules-checker tool binary.

Take a look at the following files to know the versions of those tools to use:

Using the tool docker image instead

A docker image wrapping the tool is delivered on quay:

quay.io/rhobs/obsctl-reloader-rules-checker:latest

This is actually the preferred way of using the tool as the image contains the promtool, pint and yamllint dependencies.

The only prerequisite before using the docker image is to have a container engine (docker, podman) installed on your computer.

This image can also be used on GitHub and GitLab continuous integrations to assess the correctness of your pull requests and merge requests on rules repositories. More on that later.

Using the binary wrapper

The obsctl-reloader-rules-checker file at the root of the repository is a wrapper of the tool binary.

It accepts the exact same arguments than the tool binary and make sure that:

  • The wrapped tool binary is built
  • promtool, pint and yamllint are installed

This allows using the tool out of the box without understanding how to build or locally work on it (see local development). However this script is not standalone and you have to clone the repository to use it.

Tool usage

This tool is used to assess the validity of rule files.
Imagine you have a repository containing rules definitions structured as follows:

my-tenant-rhobs-rules-clone/
├── .git/
├── .yamllint
├── Makefile
├── rules/
│   ├── rule1.yaml
│   ├── rule2.yaml
│   └── rule3.yaml
├── tests/
│   ├── test1.yaml
│   └── test2.yaml
└── template.yaml

In this example the repository contains the rules for the tenant named my-tenant.

As this tenant may differ a bit between prod and staging, we are gonna call the tool in such a way that it will generate a template; the template gather all the rules and allows to define the exact tenant through the TENANT parameter when instantiated.

Let's assume that you are at the root of the clone; you will have to call the tool as follows on Linux/Mac:

  • When using the tool binary:
    obsctl-reloader-rules-checker -t my-tenant -d rules -g template.yaml -T tests -y
    
  • When using the tool docker image:
    docker run -v "$(pwd):/work" -t quay.io/rhobs/obsctl-reloader-rules-checker:latest -t my-tenant -d rules -g template.yaml -T tests -y
    
    (replace docker container engine by podman if needed)

Remark that the flags are the same whether you are using the tool binary or the tool docker image; indeed, as a reminder, the tool docker image is just a wrapper of the tool binary with all the prerequisites needed by it.

Now lets explain the flags used in above example:

  • The -t flag is used to specify the rules tenant.
    The tenant is a string used by RHOBS to partition / shard data.
    As previously said, we are not yet sure of the tenant on which the rules will be uploaded and that's why we are generating a template.
    The value passed here is just used as a default value for the template TENANT parameter.
  • The -d flag locates the directory in which the rules are located.
  • The -g flag tells to generate a template and gives the path to the file to generate.
  • The -T flag gives the path to the unittests directory.
  • The -y flag tells to run yamllint on all the rule files and on all the unittests.
    Remark that the .yamllint at the root of the clone repository is telling how those YAML files should be formatted. This file is optional when using the docker image; indeed the tool docker image is bundling a default yamllint config file defined there.

Use the -h or --help flag of the tool to know more about the possible usages.

Tool checks

As previously explained, the purpose of this tool is to check the validity of the given rule files.

Once again, the -h / --help flag is pretty explicit about those checks. Here is a brief list of those checks:

  • Check that all rule files store PrometheusRule objects.
  • Check that objects have all a different name.
  • Check that the spec part of those objects are valid against promtool check rules.
  • Check that the objects names and tenant label are properly set.
  • Run all the unittests with promtool test rules.
  • Run yamllint on the rule files and the unittests.
  • Run pint on the spec part of the PrometheusRule objects.

Building the tool binary

Prerequisites:

  • You have to checkout the code.
  • The following tools must be present on your computer:
    • make
    • go (version 1.20 or later)

To build the binary you just have to run:

make build

The binary will be delivered in the bin folder. promtool & pint are also built and installed when running that command.

Building the tool docker image

Prerequisites:

  • You have to checkout the code.
  • The following tools must be present on your computer:
    • make
    • a container engine: either docker or podman

To build the binary you just have to run:

make docker-build

The docker image will be tagged as follows is your local registry:

obsctl-reloader-rules-checker:latest

Local development

Local development is not just about building the code. It is also about making sure that your changes will pass all the checks performed by the continuous integration (CI) jobs.

Prerequisites:

  • You have to checkout the code.
  • The following tools must be present on your computer:
    • make
    • go (version 1.20 or later)
    • a container engine: either docker or podman
    • yamllint (eventually run make yamllint-tool to install the tool)
    • golangci-lint

As you can see you have to install some linters on top of "building the code" prerequisites.

Rapidly checking that your change is okay. Just run:

make

Checking that your change will really pass the CI. Just run:

make pr-checks

The difference between the 2 commands are that:

  • make builds the tool binary while make pr-checks build the tool docker image.
    Building the code outside a docker image is faster as go can benefit some caching when building locally.
  • make will format the code and update go.mod and go.sum files if needed; you will need to commit those files alongside with your change.
    make pr-checks also updates those files but fails if the differs from what is committed.

Cleaning the repository clone. Just run:

make clean

This will remove the bin folder in which the tool binary has been delivered but also the .bingo folder which was used to build promtool & pint.

Contributing

You have to perform the following checks prior sending your code in a PR:

  • Run make pr-checks: this will build the code and make sure it passes all linters.
    More details on that in the previous section.
  • Make sure your commit messages follow the Conventional Commits specification as this will be checked by the CI.
    You can check that locally by installing a hook on your repository clone; see below.

Installing a hook to check commit messages:

  • Install pre-commit tool if not already done.
    This is platform specific, refer to the tool documentation for instructions.
  • Run the following command:
    pre-commit install --hook-type commit-msg
    

The hook is only installed on your repository clone. You my want to remove it when performing some advanced operations, see below instructions.

Uninstalling the hook checking commit messages:

  • Run the following command:
    pre-commit uninstall --hook-type commit-msg
    
  • Or just:
    rm .git/hooks/commit-msg
    

Delivering the code

Just tag your local clone and push the tag without going through a pull request:

git tag 1.0.0
git push [email protected]:rhobs/obsctl-reloader-rules-checker.git main --tags

This operation is reserved to repository maintainers. Also:

  • Replace 1.0.0 by your version.
  • Make sure you pulled the latests changes (i.e. commits) from the central repository

obsctl-reloader-rules-checker's People

Contributors

nikokolas3270 avatar typeid avatar bergmannf avatar raphaelbut avatar yftacherzog avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.