desdemona

[![Build Status][buildimg]][build] [![codecov.io][cimg]][cov] [build]: https://jenkins.racksec.io/job/desdemona-master/ [buildimg]: https://jenkins.racksec.io/job/desdemona-master/badge/icon [cov]: https://codecov.io/github/RackSec/desdemona?branch=master [cimg]: https://codecov.io/github/RackSec/desdemona/coverage.svg?branch=master

Experiments in data-driven security operations using stream processing.

Usage

Launch the Sample Job in Development

Run the deftests in test/desdemona/jobs/sample_job_test.clj. The tests automatically start and stop the development environment, so make sure you don't already have the dev environment (explained below) running - otherwise you'd get a port conflict.

Using from a REPL

Fire up your favorite REPL or use lein repl (ideally with rlwrap so you get reasonable line editing). You will be placed in the user namespace, which should come pre-loaded with a number of functions to help you set up a development environment, notably:

• (go 4) sets up a development environment with 4 peers.
• (reset) resets your development environment.
• (stop) stops your development environment.

You can launch a sample job as follows:

(user/go 4)
(require 'desdemona.jobs.sample-submit-job)
(desdemona.jobs.sample-submit-job/submit-job user/system)

Using Docker Compose

First, build the project:

docker-compose build

If you want to run the tests in Docker, you may now do so:

docker-compose run test

Start the cluster:

docker-compose up

Wait until it's all started. It should say this and then wait:

peer_1      | Started peers. Blocking forever.

Make sure you create the Kafka topic by connecting to the producer:

script/connect_kafka.sh

Now you can submit a job:

script/submit_job.sh

Anything you send to syslog on that Docker host (there's a syslog-ng relay running as a container) will appear in MySQL.

Debugging docker-compose up

If you're using docker-machine, your VM should have at least 2048MB RAM. This is because the Kafka container tries to pre-allocate quite a bit of memory. If it has insufficient memory, you'll see the following error message in the docker-compose output:

kafka_1     | # There is insufficient memory for the Java Runtime Environment
to continue.
kafka_1     | # Native memory allocation (mmap) failed to map 1073741824 bytes
for committing reserved memory.

You can fix this by increasing the amount of RAM in the VirtualBox VM to 2048MB or more.

Production Mode Peers

First start the Aeron media driver, which should be used in production mode, by running the main function in src/desdemona/launcher/aeron_media_driver.clj.

Then launch the src/desdemona/launcher/launch_prod_peers.clj main function, giving it an Onyx ID. Optionally, a Dockerfile has been created at the root of the project to package this up into a reproducible Java 8 environment.

Launch the Sample Job in Production

Launch the src/desdemona/launcher/submit_prod_sample_job.clj main function, giving it an Onyx ID and ZooKeeper address.

Run the front-end for development

First, make sure you have Leiningen installed and then run lein figsass from the project directory. This will start a Leiningen server at http://localhost:3449/ as well as automatically watch, compile and inject the project source files for you.

lein figsass is just an alias running two tasks in parallel: scss :dev auto and figwheel. For Sass compilation, project assumes you have sassc binary installed. You can use different binary by changing the :executable under :scss inside project.clj.

License

Copyright © 2015 Rackspace Hosting, Inc.

Distributed under the Eclipse Public License either version 1.0 or (at your option) any later version.