Load average
The load average is the average number of processes that are running or waiting to run (including those waiting for disk I/O) over the last 1, 5 and 15 minutes. uptime, 'top' and /proc/loadavg show it. It must be compared with the number of CPU cores.
1Reading the numbers
'load average: 0.50, 1.20, 2.00' means the load is falling: lower in the last minute than over 15 minutes. On a machine with 4 cores, a load of 4 means the CPUs are fully used on average; above that, processes wait.
2CPU or disk
On Linux the load also counts processes in uninterruptible sleep, usually waiting for disk or network storage. A high load with low CPU use in top often points to slow I/O; the 'wa' value in top shows time spent waiting for I/O.
3Number of cores
nproc prints the number of available processing units. A load average consistently above that number deserves a look.
How to check the load
- Show the load:
uptime - Show the number of cores:
nproc - Find what uses the CPU: top (press P to sort by CPU, M by memory)
- If CPU use is low but the load is high, look at I/O wait ('wa' in top) and disk activity.
Cautions
- A short spike in the 1-minute value is normal; look at the 15-minute value for trends.
- Compare with the core count before deciding the load is high.
Related articles
- OOM killerWhat the Linux out-of-memory killer does, how to find out which process it killed, and how to make it less likely.
- systemd units and systemctlWhat a systemd unit is, the common unit types, and how to start, stop, enable and inspect services with systemctl.
- Swap space on LinuxWhat swap is, how to see it with free and swapon, how to add a swap file, and what vm.swappiness changes.
Sources
- uptime(1) manual page man7.org
- proc(5) manual page: /proc/loadavg man7.org
Note: RainServer offers monitoring and safe cleanup for Linux servers; it shows what will change and asks first.
Last reviewed: