Once your code is deployed, you need to monitor what's going on. This requires upfront investment in figuring out the monitoring architecture and setting things up so that you are not blind once the application goes live. Without monitoring, when there are inevitable outages, your team will have limited insight into what's going on. This will hamper debuggability and ultimately impact the customer experience.
This is why proper monitoring is essential. There are five aspects to monitoring:
- Proper logging
- Proper emission of all relevant and important metrics from the application and infrastructure
- Well-designed dashboards that reflect the health of the application
- Actionable alerting system on the key metrics measured
- Having a production reliability team, which might include a Site Reliability Engineering (SRE) team and an efficient on-call rotation schedule...