Skip to content

Latest commit

 

History

History
102 lines (70 loc) · 2.75 KB

README.md

File metadata and controls

102 lines (70 loc) · 2.75 KB

Pachyderm Client for Scala

This is a Scala client for Pachyderm. It uses a modified version of the pfs.proto file to automatically generate an asyncrhonous Future-based gRPC client.

Currently the client is pretty barebones to the Protobuf-based client, but we will be adding some nice abstractions and utilities to make things like committing multiple files asynchronously easy. In addition, we will create an Akka-based version of the client that will allow for truly asynchronous IO and easy integration with Akka-stream based applications.

Installation

pachyderm-client is cross-compiled for Scala 2.11 and 2.12 and is available on Maven Central for release versions and OSS Sonatype Snapshots for snapshot versions.

Snapshot Versions

// Add OSS Sonatype Snapshots repository
resolvers += "Sonatype OSS Snapshots" at "https://oss.sonatype.org/content/repositories/snapshots"

// Add pachyderm-client dependency
libraryDependencies ++= "ml.combust.pachyderm" %% "pachyderm-client" % "0.1.0-SNAPSHOT"

Release Versions

Releases are hosted on Maven Central, which is a default SBT resolver.

// Add pachyderm-client dependency
libraryDependencies ++= "ml.combust.pachyderm" %% "pachyderm-client" % "0.1.0"

Running Tests

To test locally, follow the instruction for getting a local Pachyderm cluster running. After you have Pachyderm running locally, make sure you have port forwarding enabled. The tests will use a shell to determine the host IP for Pachyderm and uses a hardcoded port of 30650.

sbt pachyderm-client/test

Basic Usage

Make sure you can run the tests using the instructions in the previous section.

pachctl port-forward &

By default Pachyderm runs on port 30650 and the IP address of the Pachyderm daemon can be determined using:

minikube ip

Once you have the port and host IP for the Pachyderm daemon, we can connect to it using the Scala client.

import sys.process.Process
import ml.combust.pachyderm.client.Client

// Use a shell process to get the host IP for Pachyderm
val hostIp = Process("minikube ip").lineStream.head

// Use default port to connect
val port = 30650

// Connect our gRPC client to Pachyderm
val client = Client(hostIp, port)

// Create a repo
client.createRepo("my_test_repo")

// List all repos
client.listRepo()

// Inspect a repo
client.inspectRepo("my_test_repo")

// Delete a repo
client.deleteRepo("my_test_repo")

Working with Commits

TODO: Still working on nice utilities for simplifying commits

Block API

TODO: Still need to make the Block API integration