Contributing to Meshery Server

As a new contributor, you’re going to want to familiarize with the project in order to resolve the issues in the best way. Installing and playing around with Meshery will give you context for any issues that you might work on.

Once an issue has been addressed, you’ll need to test it as well. Ideally, these tests are run from the user’s perspective (someone running Meshery in a container), not from a contributor’s perspective (someone running Meshery as a locally-compiled service).

Compiling and Running Meshery server

To build and run Meshery server from source:

  1. Build the static assets for the UI by running

make ui-setup
make ui-build
  1. Build & run the server code by running

make server

Any time changes are made to the Go code, you will have to stop the server and run the above command again. Once the Meshery server is up and running, you should be able to access Meshery on your localhost on port 9081 at http://localhost:9081. One thing to note, you might NOT see the Meshery UI until the UI code is built as well. After running Meshery server, you will need to select your Cloud Provider by navigating to localhost:9081. Only then you will be able to use the Meshery UI on port 3000.

Please note: If you get error while starting the server as “Meshery Development Incompatible” then follow the below guideline 👇

Potential Solution:

  • Go to your meshery folder in your local-system where you’ve cloned it. Execute:

  • git remote add upstream https://github.com/meshery/meshery
  • git fetch upstream
  • Restart the meshery server
  • Additionally, before restarting the server, if you like to pull the latest changes, you can do: git pull upstream master

Building Docker image

To build a Docker image of Meshery, please ensure you have Docker installed to be able to build the image. Now, run the following command to build the Docker image:


make docker

Define and validate errors

Every Golang-based component within the Meshery ecosystem incorporates a utility to define and manage error messages for every error instance. This is internally done with several make commands, but one can explicitly validate with the help of the following make command. This checks and validates the errors that are present in the particular project.


make error

For more details, Error Utility

Configuring Log levels at Runtime

The server log levels can be configured at runtime by changing the env variable LOG_LEVEL defined in file server-config.env. The configuration library (viper) watches for the env file, any change in the file content results in the file_system event to be emitted and the log level is updated accordingly.

Should there be any alterations to the location or name of the environment file, it will result in the inability to configure log levels during runtime. In the event of such modifications, it is essential to update the server to preserve proper functionality.

```Available Meshery Server log levels are:

  • Panic - 0
  • Fatal - 1
  • Error - 2
  • Warn - 3
  • Info - 4
  • Debug - 5
  • Trace level - 6 ```

The default setting for the LOG_LEVEL is 4 (Info). However, if the DEBUG environmental variable is configured as TRUE, it supersedes the value set in the LOG_LEVEL environmental variable, and the logging level is then adjusted to 5(Debug).

Using custom Meshkit code for Meshery server development

  1. Checkout meshery and meshkit repository in adjacent directories.

    

$ git clone https://github.com/meshery/meshery.git
$ git clone https://github.com/meshery/meshkit.git

  2. In your meshery go.mod, update the meshkit package.

    

github.com/layer5io/meshkit => ../meshkit

    Remember to remove this go.mod change when creating pull requests.

Suggested Reading