Upon deploying your PostgreSQL databases in production, it is very important to monitor them regularly. We could categorize the monitoring of any database into three types:
- Reactive
- Proactive
- Predictive
To understand the preceding three categories better, let's consider an example of a PostgreSQL database with a size of 100 GB on a disk of size 150 GB:
- Reactive: Receive an alert when a database server crashes because it has reached its capacity of storage. This is said to be reactive.
- Proactive: You know that the database server is about to reach its threshold and add more space to avoid an outage. The impact may be very little to users if it requires a server reboot or a restart.
- Predictive: Based on the data captured over several days, you are able to predict the outage several days or months in advance. In this case, performance and availability...
