ChatOps for the Next-Gen MSP

Asser Hassan 27th March 2017

What is ChatOps?

Conversation-driven collaboration isn’t new, also known as: talk to the people around you and figure out a way to work together.

This is how humanity survived and climbed its way to the top of the foodchain.

In this digital age, ChatOps is certainly gaining traction, becoming the manifestation of conversation-driven collaboration. Technology teams around the world continue to use and create new ways of working together, controlling resources, streamlining processes, and innovating in their respective workspaces.

 

The 5 stages of ChatOps

The 5-stages analogy put forth by Atlassian’s Sean Regan pertains to space exploration history. “Not nearly as depressing as the more popular results on Google”

And it goes like this:

Sputnik:

Basic usage of a chat platform (e.g: Hipchat, Slack), lightweight business communication. Emails still dominate communication at this stage but chatting is finding its way.

Mercury:

Certain conversations or workloads start moving to persistent chat rooms, effectively replacing email threads. This creates a more conversation-centric environment and helps remote teams to collaborate, learn, share information, exchange resources, and have better visibility of what the team at large is doing.

Gemini:

Technical teams now start to use bots and simple slash commands to -in effect- make the chat platform into a control console. Integrations with internal and third party systems allow for a universal interface controlling and interacting with core business technologies.

Apollo:

This is the turning point. Elaborate usage of slash commands, integrating advanced bots with tools/platforms, allowing for workflow automation.

Chat at the Apollo stage is a multimodal platform, becoming the de-facto browser (and command line terminal) for an increasing amount of data. Entire workflows are now chat-based, the business processes (e.g: change management) have evolved to being chat-based as well. Teams switch between applications less and less, because they can query for information, take actions, or trigger automated procedures inside the chat.

Elon Musk:

As the name suggests, this is the endgame.

Automating large portions of mission critical work, advanced custom bots, deep integrations, attempts at AI.

Basically turning the chat into an operating system.

 

Next-Gen MSP, mission statement

By now the shift from on-premises data centers to hyperscale cloud solutions is a strategic pivot for an ever-increasing number of organizations. This comes with a new paradigm for MSPs (Managed Service Providers).

Today’s MSPs are expected to create high value solutions and position themselves as multi-cloud experts, capable of architecting solutions, managing the transition from legacy systems, and automating large portions of the workloads. The Next-Gen MSPs are leaders and visionaries, enabling their customers to leverage new technologies, and adopt the highest standards for tools, processes, and automations.


Challenges

A successful MSP should be equipped to handle a number of challenges, including the shift from pure ITIL-based operational model to a more dynamic agile model, incorporating elements of design, architecture, automation, CI/CD, DevOps, security, and financial optimization. The workloads are becoming increasingly dynamic and should be handled in an equally dynamic way.

Using multiple tools, having different processes for different customers, dynamic SLAs for different stacks, and a dynamic CI/CD-based change process is certainly the way to go. However, it comes with another challenge for the technical team running it; How do we use all the tools and resources we need without wasting unnecessary time switching screens.

This is where the introduction of ChatOps can be a game changer.


How we do it

 

At Cloudreach, we are proud to be in the Apollo phase of ChatOps.

Our teams use both HipChat and Slack for technical discussions, arranging changes, querying our knowledge base, and triggering chat-driven bots to carry out specific workflows.

We have automated notifications for certain types of communications, and advanced bots to carry out sophisticated technical tasks. Our bots leverage serverless architecture, and interface seamlessly with our internal tooling, monitoring and ticketing systems.

We have the capability to conveniently store dynamic information using simple keywords, once a team member starts talking about the keyword (e.g: an increase in 400 errors on a website, or a temporary change in a contact process), a bot will display all relevant information entered by another team member. The team members can trigger the bots to take corrective actions using natural language (e.g: bot fix issue #42)

 

Example:

Triaging certain types of security incidents would take an engineer around 15 minutes to identify the threat, link it to a cloud resource, take corrective action, and log the entire incident in a ticketing system.

The bot does that in 30 seconds.

The principle is: if a task will be repeated more than three times, we need to automate it. Bots are a quick and convenient way of doing so.

This is one of the ways we leverage brainpower over manpower. Ensuring that our most valuable resources, our people, are free to do the things that only humans can.

The future

 

To be truly Next-Gen means that our workflows are automated, our processes are geared towards dynamic change, and our information sharing and collaboration capabilities are maxed out.

As a Next-Gen MSP we’re making sure that most of our time is spent enabling innovation; creating new architectures and improving existing ones.

 

Final words

 

At Cloudreach Operations we’re very committed to our mantras:


“Learn everything

Automate everything

Self-serve everything”

It’s an ambitious goal, and we’re certain that with the right kind of vision, the right kind of people, and the right kind of tools, we will continue being one-step ahead on all fronts.
There are several things that enable us to do exactly that. ChatOps is one of them.