What indicators would you monitor to determine when to scale up the system?

To determine the need for system scaling, the following indicators are usually monitored:

• CPU usage: if the CPU is constantly loaded close to 100%, it signals the need for scaling.

• Memory usage (RAM): high memory load can lead to slowdowns or crashes.

• Latency: increased response delays indicate overload.

• Throughput: if the system cannot handle the growth in the number of requests.

• Errors and timeouts: an increase in errors or timeouts indicates performance issues.

• Task or message queues: an increase in queue length may indicate bottlenecks.

Monitoring these metrics allows timely decisions on horizontal (adding servers) or vertical (increasing resources) scaling.