In the previous step, we have started a three-node cluster and demonstrate high availability and fault tolerance. In this step, we are going to show developers how to debug in cluster mode. This tutorial is for advanced developers trying to debug some cluster-related issues locally, and other users don’t need to learn it. To run a cluster in a mixed-mode, we need to start node2 and node3 in docker-compose and node1 in the IDE with config/local as externalized config. Read More »

There are two ways to start local confluent platform to support local development and testing. Local Installation If you work on a Linux desktop, download and install the confluent platform is the easiest way, and all the command line tools will be available with the command line on your desktop. You can use a shell script to create/delete resources. confluent local services start Docker Compose If you have docker installed on your desktop, you can start the confluent platform locally with docker-compose. Read More »

In most situations, the light-scheduler will be running in a cluster mode. For developers, chances are you need to run the local application in cluster mode to test some use cases. To do that, we will give you the option to start three nodes from a docker-compose in the command line. Note: we cannot run multiple instances of the light-scheduler in the command line as these instances are not allowed use with the same local store. Read More »

The specification for the light-scheduler is straightforward. There are two endpoints with schedulers as the path. The get endpoint will get all the scheduled task definitions. The post endpoint will create/update/delete the task definitions. Specification You can find the spec at model-config repository. Get Endpoint curl -k --location --request GET 'https://localhost:8401/schedulers' Post Endpoint Create curl -k --location --request POST 'https://localhost:8401/schedulers' \ --header 'Content-Type: application/json' \ --data-raw '{ "host": "lightapi.net", "name": "com. Read More »

The following are steps to get the development environment ready on your desktop. Confluent Platform You can start the Confluent Platform with confluent local or confluent docker, and most users prefer the latter. https://youtu.be/L5RoJvCbkT0 Create Topics We need to create at least two topics for the task definitions and the other for the scheduled tasks. By default, we are using the light-scheduler topic name. Create the following two topics with 5 partitions from the control center. Read More »