You cannot fix what you cannot see. Knowing how to check your VPS's CPU, memory, disk and network turns "the server feels slow" into a specific, fixable cause. Here are the commands worth knowing and how to set up monitoring that warns you before things break.
The quick health check
SSH in (see Connecting to your Linux VPS over SSH) and run these.
Live overview with htop
htop
htop (install it if missing) shows CPU per core, memory use, and the processes using the most, all live. Watch for a single process pinning a core, or memory near the limit.
Memory with free
free -h
The available column is what matters, not free. Linux uses spare RAM for cache, which is normal; low available is the real warning sign.
Disk space with df
df -h
A full disk causes strange failures and crashed services. If a partition is near 100 percent, find what is using it with du -sh * in the large directories.
What is listening with ss
sudo ss -tulpn
Shows which services listen on which ports · useful for security and for spotting something that should not be exposed.
Reading the numbers
- High CPU on one core, low on others: normal for many apps and game servers, which are single-threaded. A faster per-core CPU helps; more cores do not.
- Memory
availablenear zero: add RAM or reduce what is running. Constant swapping makes everything slow. - Load average above your core count for sustained periods means the server is overloaded.
Set up ongoing monitoring
Spot checks are reactive. To catch problems early:
- A lightweight agent such as Netdata gives a real-time dashboard of every metric with almost no setup.
- Uptime monitoring (an external service that pings your site) tells you the moment it goes down, from outside your server.
- Alerts on high CPU, low disk or low memory turn monitoring into early warning instead of an autopsy.
Set a disk-space alert at 80 percent. A full disk is one of the most common and most preventable causes of an outage.
FAQ
How do I check what is using all my CPU or RAM?
Run htop and sort by CPU or memory; the top process is your culprit. For a quick non-interactive view, top -b -n1 | head -20.
My disk is full, how do I find what is taking space?
Use df -h to find the full partition, then du -sh * inside its directories to drill down to the large files or folders.
Do I need a monitoring tool, or are commands enough?
Commands are great for spot checks. For anything you need to stay online, add an agent and external uptime monitoring so you are warned automatically.
Want monitoring handled for you? Managed Vastrox plans include resource monitoring and alerts. Talk to support.