Fixathon
First, I need to set a backdrop as to what my team does and why this post was necessary.
Among other things, there are five products in my portfolio and they’re all super critical for Walmart.com. No, I’m not saying this to substantiate our existence and/or worth, but, these products indeed make or break the website. We have a mix of everything — messaging, microservices, orchestrators, several big data technologies and APM solutions. We use everything from simple bash scripts to Apache Flink (our newfound love!) and so many other things in-between. Needless to say, we’re polyglots. Discounting myself, there are 8 engineers who support these 5 products. We’re super slim on process, very critical about meetings, extremely vocal about inefficiencies and never shy to question the status quo. Trust me, it’s not easy to convince this bunch with BS.
Like any other critical function anywhere else, we are always inundated with features, capabilities to add to our platform and have a really long laundry list of things to do ourselves. My groups’ Engineering Leadership keeps raising the bar higher (which demands another post by itself!), like they should, and our pet projects keep going down the priority list.
We all agreed that we have to do something about it and fix it. Hence, this Fixathon. We decided that we will spend a full day and knock a few things out of the park — stuff that we’ve been wanting to do for a long time, but, never got around to doing it for many reasons.
We are truly democratic when it comes to such decisions (There are a few things where I descend on them to do it, but, not for something like this). So, I did a survey to gauge interest with the team and, not surprisingly, their responses were pretty positive.
As you can see, folks are never shy to say what they really feel. Also, we all converged on what we wanted to do that day.
We spent less than an hour, a few days before the Fixathon day, to see what every one of us was going to do and we cherry picked things that needed fixing, not today, but yesterday! Regardless of what we chose, we agreed that it has to drive significant impact — reduced API errors, improve response times, improve our streaming pipelines to increase throughput/remove inefficiencies, improve developer productivity and have better eyes and ears on our platform with real alerts. We all vehemently agreed that we should hide out in another building, something other than our usual one, as we don’t want to be seen by anybody! I told my boss that the entire team is hiding out (didn’t tell him where until later in the evening when he wanted to swing by).
I put up some guidelines for the day to help focus what was important that day!
While Slack has been nothing shy of amazing in our day to day work, it can get really chatty when all you want to do is get in the zone. Also, we all know how caffeine and engineers are joined at the hip, so, had to make sure that we leave no trace. In hindsight, I should’ve counted how many cups were consumed and see if it was directly proportional to the lines of code we wrote that day!
Some of our consumers and Ops teams were freaking out, as expected, when they couldn’t find anybody in their usual place nor on Slack. I, repeatedly, assured everyone that the team is safe, sound & happily fixing things. I was the only one fielding all questions/emails that the team would normally get via Outlook and Slack. Ironically, there were two fires that needed attention the same day. But, nope, I wasn’t going to let anything get in the team’s way that day!
We wanted to go on for full 24 hours, but, we wrapped up in a little over 16 hours. We accomplished pretty much whatever we set out to do that day. We can’t wait to deploy these to production in the coming days (still needs a little tweaking and operationalizing) and see the positive impact that its going to create in our platform and for consumers of our APIs. I couldn’t be more proud of what this team has accomplished!
It would be remiss of me if I didn’t share the Awesome 8 in my team.
I cannot wait for the next one already!
