Introduction

I want to first to introduce myself before I start publish some blog entries. This way you can get insight into my background and the experiences that drive the blog entries. Then afterwards I will explain a little about what type of blog entries I will be writing.

Introduction

I started out my career as an extreme blue intern, IBM top internship program back in 2007. Then I graduated from college and started out in IBM full time in 2008.

When I started at IBM one thing which I saw was a need for virtualization. Everyone was constantly reinstall and rebuilding their development environments, which took a lot of time. I started to build an innovation virtualization system as a side project, which later became my main job. This is back in 2009, where IaaS was just starting out. The virtualization system which I built is more similar to docker than a VM management or IaaS.

The virtualization system works by having the end user build a component by installing an application and all the dependency in a given directory. The component could be launched at a later time by combining an OS at launch time. The component can also be cloned for making a new image, or save in a catalog for others to use. For example some made a java web application server component. An end user makes a new instance by choosing the java web application server component. The instance then is launched. When the instance is launched it picks an OS, like Centos. Launches the OS then mount the component inside. The end user has a working java web application server with no need to install. They also don't need to manage the OS including for security compliance.

The solution can also combine multiple components together, for example a java web application server and a database server. Even though the 2 components were build separately. The components can also be saved together to build more complex applications. For example, you can make an instance with a java web application and a database. Install a web application and configure it to use the included database. Save the results as a new component. Then the next user can just provision it, and when the instance is launched they already have the application installed and configured properly. This can also be done across multiple instances by the use of IP based fencing, where each machine will always get the same IP address and hostname on an internal network. I was working on this till about 2012. I filed for 3 patents based on this system.

Then I moved over to the IBM cloud division. While I was there, I got the opportunity to work on a new project to help IBM adopt the latest techniques. This was done by starting from the ground up to figure out which are the best techniques and demonstrate the usefulness of them. Then use the team as a model to influence the rest of IBM development. This started back in 2013.

In that team we pioneered different transformation like open offices, test driven development, CI, CD, cloud deployment via PaaS, DevOps, incident learning reviews (postmortem), blue-green deploys, A/B testing. Through the demonstration it helped bring about transformation throughout IBM. The project grew and transformed into a standard project, with less on a focus of transformation. Even though the project accomplished a lot. More transformation was needed. At that time there was another group in IBM that was transforming the designing landscape in IBM. They were evaluating the development landscape as well, so I moved over to the design organization, this was in 2015.

One of the blockers which prevented teams from transforming was tooling. The tools which were provided did not support the modern practices. Teams either needed to stand up their own tools, which was very expensive or use unsupported tools. The unsupported tools may be a server running under someone desk. In the design org we formed a group to bring in first class development tooling which support the modern practices. Host the tools to a high standard using modern practices.

We brought in tools to support modern practices like GitHub, Travis and Slack. One thing that we quickly ran into is the tools couldn't scale to IBM workload. Even though we pick tools that operated at internet scale tools, they were package in a way to make it easy to install and manage, which reduces their scalability. We needed to work with the vendor to make it, so they could scale to our usage. Through this work we also participated in the SNAFUcatchers workgroup, which studied reliability. As I got deeper down this path I was moving more towards the tech leadership side and away from the practitioner. Recently been focusing more on security.

Blog entries

Now that you know a little about myself it is time to discuss what to expect in the blog entries. I am going to be focusing on bigger issues. Not the latest news story or the latest technology popping up. This is mainly because it takes me a long time to write the blog entries. This leads to a situation when I am ready to release an entry about the latest technology there will be many articles already published which will explain the technology better than I could have.