You will probably hear more and more about ChatOps - at conferences, DevOps meet-ups or simply from your co-worker at the coffee station. ChatOps is a term and concept coined by GitHub. It's about the conversation-driven development, automation, and operations.
Now the question is: why and how would I, as an ops-focused engineer, implement and use ChatOps in my organization? The next question then is: How to include my tools into the chat conversation?
Let’s begin by having a look at a use case. The Closed Looped Incidents Process (CLIP) can be rejuvenated with ChatOps. The work from the incident detection runs through monitoring until the resolution of issues in your application or infrastructure can be accelerated with improved, cross-team communication and collaboration.
Step 1: Keep people informed - post significant alerts into the chat
Dedicated chat channels for teams with a common responsibility help to focus. To keep everybody informed about critical events, you can let Hewlett Packard Enterprise Operations Manager i post selected events into the chat. This approach creates awareness in a very lightweight way. Here is an example of an OMi event in a chat.
Step 2: Act fast - enable quick analysis
You usually need to gather more information for assessing the situation. For example, validating the health of managed servers can be easily done with HPE SiteScope application monitoring software. Through the SiteScope bot you can run monitors via the chat to retrieve the actual state of the managed server. Additionally, the OMi bot offers chat commands to run OMi tools for further troubleshooting.
Step 3: Escalate and collaborate – create virtual IT war room easily
In many situations, you need experts from various teams to identify the root-cause of major incidents. Instead of lengthy, unproductive phone conferences, you use ChatOps to collaborate faster and far more efficiently. In our example use case, we escalate the event to HPE Service Manager service desk software. Within Service Manager, we then create a major incident. This action creates a new, dedicated channel – the virtual, IT war room. Service owners and subject matter experts are invited to this IT war room.
Step 4: Communicate, resolve and document on the fly
The communication between all members of the IT war room is transparent and visible to everybody. This is true even for people who are invited after some time—communication between humans but also between humans and bots is visible to all. Here is an example of how Natalie runs an OMi ping tool from the chats.
Once the incident is resolved, you can also update the record in HPE Service Manager and attach the whole chat conversation to the incident as a journal update:
ChatOps drives agility into your monitoring and incident management approach. Inducing your IT management tools with bots into the chat conversation creates a persistent audit trail and archive of searchable knowledge.
Getting started is simple. Many bots for various tools are available as open source. HPE Software bots can be downloaded on GitHub.
Visit the ChatOps home page on Live Network to find information about HPE Software ChatOps and bots.