Post

One of the less talked-about aspects of serverless is the code rot that can occur strictly based on the fact that the services can happily run for literal years unattended. Not even talking just dependency updates, I just mean organizationally you can just lose tribal knowledge entirely around it.

0 reposts10 likes

That’s the same for serverful environments. I remember finding random services running on servers that nobody in the team recognises anymore and they have been running for years. I’m sure they are still there… no one dares to touch these…

Sure. I’m not saying this is exclusively a serverless thing, or even a problem! It’s just lot easier thing to happen when all the underlying stuff is patched and tolerates outages and even memory crashes, etc. if you aren’t actively dragging along those services they won’t evolve as your org does.