Hacker News Reader

Comment by kovac | Hacker News Reader

I did read the article. Saying your now small team to be "careful" is not a real solution which I mentioned in my comment. What do you mean by expressing my superiority? Expressing my opinion backed by my own professional experience which happened to disagree with your own?

View on HN

Incorrect.

Eloquent. So... we should change application architectures as the team changes size?

That appears to be what the article is suggesting.

Team size and structure definitely is relevant, as many of the main arguments for or against specific architectures are about how it makes certain developer processes faster or safer or easier or more predictable; so yes, it seems reasonable that the most appropriate architecture would change as the team changes size - though it does not mean that we should change the architecture, as the benefit may not be worth the cost of change.

Changing a system architecture twice like this, specially to go back to what it was before is a large cost to a company. In this line of thinking, they may have to move back to a microservice if they had a bigger team for the exact same application! This is fine by me, but rather than what other benefits they enjoyed by migrating the second time, a clear analysis of motivations behind moving to microservice and then back to monolith without compromising the benefits attained by doing the first migration would be a lot more enlightening. Because I don't imagine the first migration to microservices happened just because they had too big a team. I suppose this thread will naturally attract people who hate microservices in principle.

Logic:

Because of microservices, we had bugs (okay, what kind of bugs?). Therefore we moved back to a monolith again (why did they use a microservice architecture in the first place). Then the bugs disappeared (how?) and we got a whole bunch of other benefits as a bonus (any trade offs?).

Cool, so the initial migration to microservices must have been done by complete idiots then. It seems that even to question that is too much here.

I think you're starting to get the point:

There are idiots out there. And they're doing microservices. For no reason besides hype.

The article says exactly why they moved to microservices: too many teams were deploying the monolith and it slowed them down.

The article also says why they are moving back to a monolith: the services in question are legacy and being replaced, and only one team is maintaining them, so the reason they moved to microservices no longer applies. In addition, combining the services makes their lives easier enough that it’s worth it from their perspective.

Your aggressive tone and assumption that they didn’t have good reasons for what they’re doing is why I said you were “rushing to express superiority.”

Reply on news.ycombinator.com