Athena

This is a fully asynchronous and nonblocking Cassandra client built on top of the (Excellent) Akka-IO framework.

Currently, it can create a fault tolerant pool of load balanced connections to a Cassandra cluster, stream queries and marshal results.

What doesn't work yet -

Authentication
Connection compression
Intelligent (key aware) query routing
Documentation - sample code

Using Athena

Binary release artifacts for Scala 2.11.x and 2.10.x are published to the Sonatype Maven Repository and synced to Maven Central. The current release is only available as a SNAPSHOT (for now), but the final 0.3.0 release is imminent.

SBT builds

libraryDependencies ++= Seq(
  "com.vast" %% "athena" % "0.3.0-SNAPSHOT"
)

Maven builds

<properties>
    <scala.major.version>2.10</scala.major.version>
</properties>

<dependencies>
    <dependency>
        <groupId>com.vast</groupId>
        <artifactId>athena_${scala.major.version}</artifactId>
        <version>0.3.0-SNAPSHOT</version>
    </dependency>
</dependencies>

Usage

TODO: More info here

Actors

The high level interface for Athena is through the ClusterConnector actor. This actor models a pool of connections to all the hosts in your cluster, and will load balance queries across them in an appropriate fashion.

Sessions

There is currently an experimental higher level Session abstraction that models interactions as a series of pipelines, where a pipeline is simply defined as

AthenaRequest => Future[AthenaResponse]

A Session wraps a pipeline and adds more useful methods on top of it, and also the ability to stream rows back using Play Enumerators. The entirety of this API, however, is really just a prototype and is subject to change across versions with little to no warning.

Building

Athena is built with SBT 0.13.x, and is designed to be essentially and out of the box build. You should be able to compile, package and (if needed) deploy locally by simply doing 'sbt publishLocal' from the command line. The test suite requires an active instance of cassandra to run against. Fortunately, the SBT harness is built in such a way that it automatically handles downloading the proper version of cassandra, configuring it, starting the server and setting up a test keyspace before the tests run. This instance is then torn down after tests complete. Most of this functionality is configured using SBT settings - look in project/CassandraUtils.scala for info on how to change it. For example, if you wish to disable the server auto-provisioning to use your own instance, just add

import CassandraUtils.Keys._

provisionServer := false

to your local.sbt file.

vast-engineering/athena