Mudrex’s Path to DevOps Excellence and Platform Engineering

What is DevOps?
DevOps combines cultural philosophies, practices, and tools that increase an organization’s ability to deliver applications and services at high velocity. In layman’s terms, this means, enabling developers to ship features faster and more efficiently.

In this paradigm, development teams (backend and frontend) and operations teams (infrastructure, SRE) are merged into one so that, the engineers work across the entire application lifecycle, from development to tests, to operations, and develop a range of skills not limited to a single function.

What is Platform Engineering?
Platform Engineering is a relatively newer concept in the market, which aims at tackling the problem of a learning curve in development teams. Platform Engineering aims to provide an automated interface for infrastructure operations which in turn increases the developer productivity and delivery. This also eliminates the need to have a dedicated DevOps person per team. Instead, a unified horizontal interface is made available to organizations for them to able to test, deploy and deliver without ever having the need to know what happens on the other side of the interface.

Mudrex’s Platform Engineering Team
We at Mudrex also have a dedicated Platform Team, aiming to achieve the things mentioned above. We build tools to automate mundane tasks, provide feedback loops so that developers respond faster, and maintain cloud infrastructure that keeps Mudrex running.

The actual state of Platform Engineering at Mudrex
To be honest, we are very far from an ideal DevOps state in Mudrex, but we are getting there. In the past 6 months, we have been able to stabilize infrastructure to the extent that we have not had an Infra Incident for the past 6months and we have slept peacefully knowing that if anything breaks, it gets healed automatically. But on the DevOps side, we have been lagging far behind. To this date, developers are dependent on the Platform Team to get a new service up and running, which in an ideal case developers should be able to do themselves. The current status introduces a lot of friction, delayed deadlines, and gaps when it comes to deploying something new, the whole JIRA fiasco, and the Platform Team’s availability.

The ideal state
This year’s goal has been tailored around fixing these problems. We are working to educate developers, enabling them to get some jobs done themselves without having to depend on the Platform Team, we have been building automation tools that ease the work and lower the learning curve and are easy to adopt.

An ideal state would be: Platform Team has 0 external JIRA’s and sleeps at 10 PM to get a full 12 hr sleep 😛

Some tools that have made developer’s life easier
1. Athena: Making releases as easy as typing a message on Slack. All the heavy lifting is done behind the scenes.
2. Watcher: You want to scale something, watcher does it. No questions asked.
3. Nightswatch: Observability at scale with reliability.
4. Aletheia: Central RBAC (based on LDAP) for the singular source of identity truth.
and a lot more.

We are always looking for developers to contribute to Platform Tasks. Please do reach out if you have the bandwidth in your sprint.


The latest from the blog