Thursday 

Room 4 

11:40 - 12:40 

(UTC+01

Talk (60 min)

ChatOps: Why you should give your application a voice

What if your application could send a message to the right engineer if there was something that needed attention? What if it could join relevant instant messaging channels during a production incident, responding to questions and grabbing logs you need to be looking at?

This is "ChatOps". ChatOps is a collaboration model that connects people, tools, processes, and automation into a transparent workflow. It's the bridge that connects the work that is needed, what is currently running, as well as what has been done - utilising real time communication channels to keep you updated every step of the way. It started life at GitHub, where they created Hubot and found it brought them huge efficiency benefits.

Collaboration is critical in DevOps. According to Gartner [1], 75% of DevOps initiatives fail "due to issues around organisational learning and change". ChatOps addresses that through a focus on collaboration and information sharing. This is why we have it as a key practice in our Garage Methodology [2]. It's also why we see many other businesses turning to ChatOps as a more collaborative and efficient way to perform operations. Instead of just using a traditional help ticket tool, DevOps experts use instant messaging collaboration tools like Slack to communicate - not only with each other, but also with the tools they use to do their jobs.

In this session, we want to get you thinking about whether adopting elements of ChatOps could improve your DevOps workflow and give you practical tips to get started.

To help give you some ideas for what you can do with your own application, we'll show you a demo using Apache Kafka - it's a nice simple example of a microservices application, plus it already has instrumentation and monitoring hooks. We'll show how to get a Kafka cluster to post messages to your team Slack channel when there is something that a dev-ops person should be be looking at. And we'll give an example of how to perform simple Kafka admin tasks from Slack.

[1] - https://www.gartner.com/smarterwithgartner/the-secret-to-devops-success

[2] - https://www.ibm.com/garage/method/practices/manage/chatops

Lori French

Software Engineer, Designer & Inventor working in the IBM Client Engineering Team, based in London. I have a passion for delivering innovative prototypes for Enterprises, being a member of the Wimbledon Technical Team.

Dale Lane

Dale is a developer for IBM. He spent several years as a developer in IBM Watson, helping to create several Watson cloud services. He is the author of "Machine Learning for Kids", and the creator of the supporting website MachineLearningForKids.co.uk which is used by children around the world to learn about artificial intelligence by creating and playing with their own machine learning models.