Skip to content

Latest commit

 

History

History
69 lines (48 loc) · 3.15 KB

README.md

File metadata and controls

69 lines (48 loc) · 3.15 KB

Piper

Keep in mind that this project is still only in a POC state

Piper is an early-stage tool designed for building pipelines consisting of operators that execute user-defined scripts. While still in development, the tool aims to provide a structured approach for creating and managing pipelines.

The core concept is to enable the execution of any type of operator that, under the hood, utilizes a Kubernetes resource. These operators can be executed in a specific sequence defined by the user.

To simplify the pipeline creation process, Piper employs a Kotlin DSL (Domain-Specific Language). This DSL makes it easy for users to define pipelines intuitively and efficiently.

An example pipeline can be found in the examples/script.piper.kts file. This demonstrates how Piper facilitates the orchestration of multiple operators in a clean and user-friendly syntax. Piper strictly detects files with the .piper.kts extension and supports the use of dependencies from Piper’s core library.

For instance, you can declare and include a Python script (e.g., script.py) or scripts in other languages such as Bash or KScript.

Below is a brief demonstration to serve as documentation:

Declare for instance python script called script.py or any other language (eg. bash, kscript,...)

print('[Python] Running operator 1...')

Declare the path of your scripts, containing:

val script1 = Path.of("examples/scripts/script1.py")
val script2 = Path.of("examples/scripts/script2.sh")
val script3 = Path.of("examples/scripts/script3.sh")
val script4 = Path.of("examples/scripts/script4.sh")
val script5 = Path.of("examples/scripts/script5.sh")
val script6 = Path.of("examples/scripts/script6.kts")

Declare the operators you would like to run:

val op1 = KubernetesJobOperator("first", "python", listOf("python"), listOf("/scripts/script1.py"), script1)
val op2 = KubernetesJobOperator("second", "ubuntu:latest", listOf("/bin/sh"), listOf("/scripts/script2.sh"), script2)
val op3 = KubernetesJobOperator("third", "ubuntu:latest", listOf("/bin/sh"), listOf("/scripts/script3.sh"), script3)
val op4 = KubernetesJobOperator("fourth", "ubuntu:latest", listOf("/bin/sh"), listOf("/scripts/script4.sh"), script4)
val op5 = KubernetesJobOperator("fifth", "ubuntu:latest", listOf("/bin/sh"), listOf("/scripts/script5.sh"), script5)
val op6 = KubernetesJobOperator("sixth", "kscripting/kscript", listOf("kscript"), listOf("/scripts/script6.kts"), script6)

Declare the specification of your pipeline using the library of piper:

val pipeline = pipeline {
    specify {
        op1 pipe listOf(op2 pipe op3, op4) pipe op5 pipe op6
    }
}

Declare that this pipeline should be ran and visualized.

pipeline.visualize()
pipeline.run()

Now since this script has already been defined here examples/script.piper.kts we can just run host/src/main/kotlin/Piper.kt with an argument on the filepath examples/script.piper.kts.

img.png

Once you run it, you should be able to see that the KubernetesJobOperators you've defined are running. You can use a tool like kubectl or k9s to view the execution of each job/pod.