My Docker trauma episode and where it will take us.
Ben Nadel well, I’ve started to hear your podcast chapter regarding the “interviewing” and heard you talking about your frustration with the dockers. In 2017 I tried working with Dockers over local windows machine and it was so devastating, and I was so traumatized that when we started our startup, 3 years ago, I went for the leanest version that did not need any management – azure app services. as our Backend was the .Net Core, it was just a natural choice to go with Microsoft and we event got the BizSpark freebie for startups for a full year…
Now the team starts to talk about dockers and Kubernetes but the old trauma just not allowing me yet to go there..
Also, I always feel that diving in the Kubernetes its a bit like a swamp, and without a good guide, we are going to drown there… It like the only way in, and no way out.. Another vivid association for Kubernetes I have is a spider web as I see it. And the most alarming is that even experts never really can know it to the last bit, because of the specifications and integrations of each and every case implemented.
For now we are with the app services, and by the way, going for the Disaster Recovery tests scenarios now that will map the actual resilience and scaling issues if any on the way..
So, if I refer to an issue you had with the Docker, you were saying that the Docker just died without any understandable reason, I don’t know what you tried to achieve, but maybe simplifying it and not using VM or Docker and just throwing the code as Functions and AppServices in the cloud can do a quicker job, outcome-wise, because it’s removing the need for you to be a DevOps for the project just by using already existing simple solutions.
But hey, maybe it was not your case at all…
When we got to a tech meeting with big companies most of them envy the simplicity and ability of our startup to move quickly and adjust for the architecture is lean and simple from one side, but stands its own ground in the production from the other side, overcoming performance spikes by scaling out and up whenever needed automatically but keeping it easily manageable and very cost-effective.
I guess this cry out of the heart will be the base for my next post in my blog ..
I am putting this link into my blog to other post called – Edx Sanbox – get it running on local Windows machine, that can articulate the nightmare you made me remember today.
At 2017 it took me about few days of personal time, because it was just a side project, to create a running EDX machine on localhost with Windows.
So the conclusion is, it will take some convincing for me to move, although the reality maybe will be stronger then my trauma at some point.