Eric Fleming on Middle-of-the-Day Deployments - Episode 41

Today's episode is all about recognizing middle-of-the-day deployments; how teams such as Netflix, Facebook, and even the Azure DevOps Product Team are doing them; and taking a look at how other teams can achieve that for themselves!

Jeffrey Palermo's guest today is Eric Fleming, a Software Architect at Clear Measure. Eric leads an intense team, developing and operating a mission-critical software system in the financial sector. He lives in Alpharetta, Georgia, and is a host of the Function Junction Youtube Channel, which is all about Azure functions. He's also written articles for MSDN Magazine and CODE Magazine.

In this episode, Eric takes Jeffrey through his journey of inheriting a monolithic software system and the major transformations he had to execute to get it where it is today; deploying in the middle of the day! He explains the key steps he took in breaking up the monolith, the development process, who was involved, what the structure and DevOps environments looked like, and all of the details you need to know if you're finding yourself in a similar situation!

Topics of Discussion:

[:52] How to get your hands on Jeffrey's book, .NET DevOps for Azure.

[2:04] About today's episode and featured guest.

[3:00] Jeffrey welcomes Eric to the podcast!

[3:06] Eric begins the story of how he inherited a software system and the journey it took getting it to deploy in the middle of the day.

[9:58] Fast forward to today, what does this software system look like now?

[11:50] What does Eric attribute to his ability to handle a high-throughput in only four app servers?

[15:52] Eric's process for deploying the 50-sum processes that need to be deployed.

[17:32] A word from Azure DevOps sponsor: Clear Measure.

[17:59] Eric speaks about their Git Repositories.

[19:25] Eric explains what the structure and DevOps environments of one of his applications looks like (that is a Windows service with its own Git Repository).

[21:45] Who is involved whenever part of the system is being deployed?

[25:37] Has there been development process differences during their monthly deployments/monolith time?

[26:22] Now that they are shipping every day/whenever they need to, what has become of their sprints? And how do they get some features done in just a day and ready to deploy within days? What does this look like and how do they implement this pattern?

[31:50] Do sprints even exist in this new world?

[33:31] The major transformations that Eric had to execute to get to where he is today with the software system, and some of the first steps he took to breaking up the monolith.

[36:27] Would Eric have been able to start breaking the monolith apart if he didn't have automated tests?

[38:47] Resources Eric recommends to listeners in a similar situation to where he was!

Mentioned in this Episode:

Azure DevOps

Clear Measure (Sponsor) — Reach out to Jeffrey @JeffreyPalermo on Twitter if you have a user group or conference and would like some free copies of .NET DevOps for Azure!

.NET DevOps for Azure, by Jeffrey Palermo

bit.ly/dotnetdevopsproject — Visit for an example of .Net DevOps for Azure

Function Junction Youtube Channel

MSDN Magazine

CODE Magazine
Eric Flemming's Twitter: @EFleming18

NServiceBus

Particular Software
TeamCity

Octopus Deploy

Sumo Logic

New Relic

Accelerate: The Science of Lean Software and DevOps: Building and Scaling High Performing Technology Organizations, by Nicole Forsgren PhD, Jez Humble, and Gene Kim

The DevOps Handbook: How to Create World-Class Agility, Reliability, and Security in Technology Organizations, by Gene Kim, Patrick Debois, John Willis, and Jez Humble