A Work in Progress


Development Complexity vs. Operational Complexity

I was discussing some solutions with my co-workers recently and I was advocating for one solution while some of my coworkers were advocating for other solutions. The debate was good natured, but everybody felt strongly about their opinions.

While the technical details of the proposed solutions were very different, I also realized another key difference between the solution I was proposing and one of the solutions that my coworker was fighting hardest for. I was advocating for a solution that required development complexity. It would be hard to maintain for a while (for reasons having to do with legacy code and backwards compatibility) and we would have to test it thoroughly. The other solution was less challenging technically, but would require a lot more complexity from an operational standpoint. Streaming data from one database to another, making all the data in a DB that used to bewritableread-only, monitoring the replication of data to make sure nothing goes wrong, etc, etc.

I realized that given the alternative between development and operational complexity I will almost always choose development complexity. Here are a few reasons why:

In the end, we hashed everything and came to a conclusion and we all moved forward but it was interesting to see which people tend to favor development complexity over operational complexity. As a developer, what do you favor? why? My guess is that any developer who is into the DevOps frame of mind would much rather development complexity for the reasons I listed above. Developers working with a traditional IT organization maybe more apt to throw it over the wall to IT and let them manage it.


email icon github icon medium icon stack-overflow icon twitter icon