Your control panel shows CPU, memory, disk and network over time. Reading them well tells you whether you have a capacity problem, a code problem, or no problem at all.
CPU. Brief spikes are normal and healthy - it means the machine is being used. Sustained high load, especially load average above your core count for long periods, means requests are queueing.
Memory. Linux deliberately uses free memory for disk caching, so high memory use is not itself a warning sign. The number to watch is swap. Once the server is swapping steadily, everything becomes slow in a way that looks like a mystery until you notice it.
Disk. Watch both space and I/O wait. A full disk breaks services in confusing ways - databases stop accepting writes, logs stop, and errors become misleading. Set yourself an alert at 80 percent rather than discovering it at 100.
Network. Sudden sustained outbound traffic with no explanation is worth investigating. It is sometimes a backup job, and sometimes a compromised machine sending spam.
The most useful habit is to look at the graphs when everything is fine. Knowing what normal looks like for your server is what makes an abnormal graph obvious later.
