[fsociety]

Features have been breaking in production, journalists have commented on “weird behaviors” in the app AKA consistency issues. I think long-term these small breakages will burn the engineering team out, in addition to the unnecessary drama they have been brought into.

More interestingly, I have a theory that the app has not had a large-scale breakage precisely because of the reason they had bloat.

Having a large org run on microservices makes it likely that you are running more infra than you require. But then failures tend to be self-contained in systems. This then means that on average, a failure will only affect one small part of the system rather than the whole.

It will be interesting to hear post-mortems from engineers once the spotlight dies down and people are ready to talk about what is happening internally.