I have finally finished reading “Antifragile” by Nassim Taleb and while this is a great book in general, a lot of things about Via Negativa resonated with me in regards to the DevOps field specifically.
In short, Via Negativa is a way to achieve things by not doing something, rather than by doing. Below are my thoughts how this principle could be applied to DevOps:
- Pick a lighter solution if it fits the needs. In example, for a development cluster we frequently recommend a single K3s VM instead of something much larger like Elastic Kubernetes Service.
- Novelty Budget (sometimes called Innovation Budget) – this is where you put a hard limit on the amount of new technology that may be introduced in a specific time period.
- Minimize the number of moving pieces in the stack. Yes, you can put 5 proxies – one on top of each other – but better not. Frankly, I frequently feel that untangling all the hoops of over-engineering takes more time than developing a solution in the first place. This is a sentiment that is often shared in the industry in general – when developers prefer to just rewrite the whole project from the scratch, rather than trying to update what they have.
- Minimize the number of scheduled meetings. Tons of material were written on this subject. In short, my point here goes as following: all meetings should be driven by the actual need to discuss something and then by the shape of the project. If there is a major issue with the project then a daily status meeting may be warranted, otherwise once a week all-team status meeting should be more than enough.
- Remote work all the way – cut the needless routine and commute to get to the office. I was talking about this before Covid, and what surprises me now is when some organizations still try to get everyone back to office. Seriously? Now, when we have an undeniable proof that remote work is as productive as onsite, somebody is still trying to find a way back. Hopefully, the strained job market will correct this for good.
The above is by no means an exhaustive list. Likely many more ideas and examples of what we can avoid doing to achieve better results may be added. But I hope this list illustrates the Via Negativa principle in DevOps well.