In this video:
Running continously available cloud services is a tough job (not to mention other hideous adjectives like manual, repetitive, and boring). There are lots of variables you have to deal with when you’re operating a cloud-managed application – which environment was that in, what server broke, is the network up?
IBM Distinguished Engineer Dan Berg and IBM Fellow Jason McGee (and lots of other programmers, designers, and engineers) decided “that’s where chatbots come in.” That’s how to pull together all the information to eliminate the drudgery of the task.
Continuous + Conversational ÷ the Power of Automation & Chatbots = ChatOps
What is ChatOps?
In the last decade-and-a-half, the demands of keeping up the operation of global services (with billions of users) have almost driven software and IT teams insane. On human power terms alone, it is not scalable. Things had to change, so communication processes evolved from email to chat, many admin tasks morphed from manual to automated, and service delivery mechanisms transformed from periodic change control sessions to continuous delivery (or DevOps, as it is known).
Ta-dah … ChatOps: A collaboration model that connects the people, the tools, and the processes in an automated, transparent workflow.
- Actual conversation becomes the driving force that accelerates the ability to handle such large workloads, allowing operators to match the speed of system
- Automation removes the repetitive tasks, allowing the human component to slow down to a comfortable enough pace at which the operators can make the creative decisions more effectively
- Transparency enhances the ability of anyone on the team to contribute to any part of the solution
- The flow part means all the bits and pieces of the process are connected
In this video, Dan and Jason demonstrate how chatbots combined with Watson AI (and some extreme automation) can help eliminate manual tasks for site reliability engineers. Since each deployed bot has its own persona and assigned tasks and can communicate not only with each other but with the humans as well, it is like having an army of administrators, each with his or her assigned tasks, at your call. And the bot part of this workforce doesn’t need to sleep or eat – they are 24/7.
You’ll see how bots can work together to communicate potential security threats and unusual system patterns, plus you get to watch a recorded conversation with a chatbot using Watson services.
Resources for you
- IBM Code: Create an investment management chatbot
- Learn to deploy a distributed GitLab leveraging Kubernetes™ and Docker
- Chat with the container services team on Slack
- ChatOps in action: Training Watson NLC with a web crawler
- Learn how to help DevOps teams respond faster with smarter alerts
- Look at the Hubot-style Cloudbot project, available on Bluemix
- Try IBM Cloud free