> 8. Avoid hero culture. Build systems that don’t require heroes.
This right here is the what I always have encountered at every team I have ever worked. There will be a production outage, one person comes in and be like "Oh, let me do this.." Magically fixes things that was never documented becomes a hero and then we do a root cause analysis and then never address the fact that, what that one person knew that service that no one else on the team did.
- I have nodejs server for the APIs and its running on m5.xlarge instance. I haven't done much research on what instance type should I go for. I looked up and it seems like c5n.xlarge(mentioned in the article) is meant compute optimized. That cost difference isn't much between m5.xlarge and c5n.xlarge. So, I'm assuming that switching to c5 instance would be better, right?
- Does having ngnix handle the request is better option here? And setup reverse proxy for NodeJS? I'm thinking of taking small steps on scaling an existing framework.
It says MySQL 8 and Elasticsearch 7.8. I don't have much experience in elasticsearch, I wanted to know how does elasticsearch makes it faster? Is it like an extension that makes it faster? Or Elasticsearch has its own data store that consumes data from the database and magically makes it faster?
This right here is the what I always have encountered at every team I have ever worked. There will be a production outage, one person comes in and be like "Oh, let me do this.." Magically fixes things that was never documented becomes a hero and then we do a root cause analysis and then never address the fact that, what that one person knew that service that no one else on the team did.