The letter from LizardFS’ CTO

Dear Community,

My name is Marcin Konarski, and I am a new CTO of the company that develops the LizardFS software. The announcement of the new release candidate (3.13-rc3) for the project is a great opportunity to reach out to our Community and (re)introduce the project, the team, and the company.

In this post, I will also try to explain how we all got here.

Let me first introduce myself.

I was 9 when I wrote my first program ever, I was using BASIC on Commodore 64. Right now, I have well over 2 decades of commercial experience in the IT field.

My technical interests include C++, Unix/Linux, programming language design and implementation, compilers and compiler optimization, machine learning, systems design, and architecture.

I joined the project in early 2014 shortly after the project was conceived by Piotr Modrzyk, Marcin Sulikowski, and Aleksander Lewandowski in 2013. At that time I was the one responsible for the implementation of the prototype of the shadow server.

The project.

The project started in 2013 as a fork of MooseFS by one DevOps engineer with a business vision and two C++ programmers (probably one of the best software engineers in Poland). They found business assistance and the initial funding necessary to kick start the project. The newly created company found its first client that was willing to fund the development of specific features for LizardFS. (I joined the team at that time.)

Unfortunately, there was a disagreement between the engineering and business parts of the company on how the project should be managed. The company also struggled to secure reliable funding for the project, which eventually led the whole core development team to leave the company in early 2015.Me and Marcin Sulikowski continued working on LizardFS under a different brand, while the original company hired new engineering staff and kept developing the original LizardFS version.

The company.

In early 2020 I was contacted by someone new from Skytechnology (the original company that kick-started the project) and I was invited to have a friendly chat about both the company and the project. At that time I was involved with a completely unrelated project with another firm.

During the following meetings, I have learned that the development of LizardFS had been on hold for many months. I have also learned that the previous majority stakeholder sold all of his shares to a new investor and that the whole previous business team and the previous company owner left for good.

The new sole owner of Skytechnology, a business angel with a strong engineering background in data storage, saw the potential of LizardFS software and didn’t want several years worth of development efforts to go to waste so he reached out to the original technology founders to reinvigorate the project.

And that is how I came back on board of the project. This time with both experience and power to shape LizardFS’ future.

The team.

When I joined Skytechnology (then soon to be rebranded as simply LizardFS company) again (in April of 2020) the team consisted of three young, enthusiastic, and super smart C++ developers. Unfortunately, all of them lacked the necessary experience required to set a direction for the development of a complex high stakes project.

That became my role.

One of my first pushes was to at least try to get original engineering founders on board again. And to a lesser extent, we succeeded. First, Piotr `aNeutrino` Modrzyk, the original visionary from whose mind the LizardFS project was actually born, joined the team. We are also in contact with Macin Sulikowski who is willing to help us with getting the project back on track.

(At this moment LizardFS has six experienced engineers as a core technical staff.)

The project again.

I believe that you cannot fix or improve what you cannot measure. When the development of the project was finally restarted in April of 2020, all our team efforts went into building testing infrastructure both in the hardware and in the software.

My goal was (and still is) for LizardFS test suites to run as often as possible and to have a feedback loop for the developer as short as possible.

Since we started running our test suites again (when the project was dead no one bothered to run any tests) we sped up its execution by more than 500%.

With a short feedback loop and trustworthy tests results, we can start working on fixing bugs and the development of new features.

Me again.

Why did I join a dead project? Well, I know first hand what the true value of LizardFS is. Right now it is a little known, all but forgotten gem of software. But the technology within it can actually make an impact in any business depending on SDS solutions.

It is trivially simple to configure, it is fast and reliable. It is all there, it just requires a little polish to truly shine and make a difference between impossible and viable for any project/company/deployment requiring efficient, seamless, high capacity and highly available storage on the cheap.

So I joined this project because I share the original engineering founders’ vision and belief and I see a great potential just waiting to be realized in the LizardFS project.

And I want to be the one who materializes it.

The Community.

During all of the years of LizardFS development, a lot of people outside of the core team got involved in the project and spent a significant amount of effort to improve it.

We (the core team) see you and we deeply appreciate your involvement in the project.

We will do our best to listen to all your voices and let you know that your voices were actually heard.


Marcin Konarski

LizardFS Team