What's your plan when app slows down?
Logging into the server? Developers use SSH on your production machine. You don’t know what they do there. One command miss-type or bad actor is enough to leak data or cause downtime.
Searching logs with grep? If a bug was reported couple days ago, log rotation may have deleted the logs. What if the app runs on multiple servers? It becomes too hard, and the dev might just say "can't reproduce" and wait for the next unlucky customer.
Alerts, insights, and speed metrics are only possible with centralized logs and traces.
Check it out: https://otelic.com/