A few days ago, somebody shared the image above on Linkedin. This image perfectly represents a situation that we’ve been seeing in one of our clients for years. And, of course, it represents the current status of many other companies. I’m not going to mention their names obviously, but I hope they will read it so they will reflect and do something about it. They will know who they are.
This client is a media company. They’ve been providing a platform for video streaming for years. They’ve been using a legacy method to manage their infrastructure for a long time, which didn’t represent a problem when they had 10 servers. One day a big company bought them to use and expand their user base, and the server count increased to 30. They were expecting to grow in the future. We had meetings about this and how to be prepared for it, but they were always doing firefighting and they never had time to improve their processes.
We hadn’t talked to them for about a year, when a few days ago they called us. The first thing they mentioned was “we have 150 servers now and growing”. It’s important to say they are managing this new volume using the same methods. They replaced some part of the infrastructure with cloud servers, but made no other key changes to the process.
We’ve visited them a lot of times over the past years, working on small projects, and there was something that always called our attention. They were (and they are) doing firefighting all the time: stress, people running, and a lot of unhealthy and risky situations for the people and for the business. I would need more than a simple article to describe the problems, but in short: no version control, no configuration management, poor metrics, and of course, there wasn’t any DevOps process or tool around.
The requirement they have now is to provide a tool to improve how they distribute the content to the web servers. The current method takes a lot of time, it connects to the server one by one, and there is no clear way to check the results or to rollback if something goes wrong. For us, this represents a great opportunity to help them to improve not only this process, but also to set up a base they can use to improve other processes in the short-term as well.
Recently I’ve been reading a book about the “Lean Change” method. There is an interesting concept in this book called “Minimum Valuable Change” or MVC. When you try to make a big change, you can’t do it all at once. The first step is to decide where the emergency is, and if there are many of them, to define priorities. You have to start with a small change, demonstrate the results and then move forward to use the right processes, the culture, and the tools to manage a company in 2015 with speed and agility.
Changing an organization is no easy feat, but we’d like to take on this challenge. They need help in order to change, but they should be the key players who make it happen. It’s a hard step for sure, but the technical aspect is relatively easy. The human facet requires time, and good communication. It requires creating a feedback loop between all the areas involved. Step by step is the right process to follow.
We hope to be writing another article about the positive results of this soon.