Load average shows how busy a system has been, with three numbers representing the average load over 1 minute, 5 minutes, and 15 minutes. These values indicate how many processes were using or waiting for CPU resources. A load of 1 means one process was active; higher values depend on core count—on a 4-core system, load under 4 is normal, around 4 is fully utilized, and consistently above 4 is overloaded. The 1-minute value shows immediate issues, the 5-minute smooths recent trends, and the 15-minute reveals long-term patterns. High load can result from background jobs, traffic spikes, or stuck processes. If CPU usage is low but load is high, check disk I/O or swapping. Use tools like top, htop, or uptime, monitor historical trends, set alerts based on core count, avoid reacting to short spikes, and combine load data with other metrics for better diagnosis.

When you see "load average" on a system monitor, it's not just a random metric—it tells you how busy your system has been over different time intervals. The three numbers represent the average system load over the past 1 minute, 5 minutes, and 15 minutes respectively. These aren't percentages; they’re averages that reflect how many processes were actively using or waiting for CPU resources during those periods.

What Exactly Is Load Average?

The load average is a measure of how much work your system is handling. It shows the average number of processes that are either running or waiting to run (for CPU time) over specific time windows. This includes processes using the CPU, disk, or any other resource that causes them to wait. So even if your CPU isn’t maxed out, high disk I/O or blocked processes can push the load up.

For example:

• If your load average is 1, it means one process was using or waiting for the CPU at any given moment.
• If it’s 4, then four such processes were active on average—this could mean full utilization on a single-core machine or moderate load on a multi-core system.

How to Interpret the Three Numbers

The three numbers in the load average give you trend data:

• The first number (1-minute average): shows what’s happening right now. A sudden spike here might indicate an immediate issue.
• The second number (5-minute average): smooths things out a bit, showing whether the system is still recovering from a recent workload.
• The third number (15-minute average): gives the big picture. It helps determine if the system has been consistently overloaded or just had a temporary blip.

A good rule of thumb: compare these values to the number of CPU cores in your system. For example, on a 4-core machine:

• Load under 4: normal operation
• Load around 4: system is fully utilized
• Load consistently above 4: system is overloaded

When Load Average Matters Most

You should pay attention to load average when diagnosing performance issues. High load can cause sluggishness, timeouts, or even crashes if left unchecked. Common situations where load becomes relevant include:

• Running too many background jobs at once
• Dealing with sudden traffic spikes on a web server
• Processes stuck in a loop or waiting endlessly for disk or network responses

If you're seeing high load but low CPU usage, check disk I/O or swapping. Tools like top, htop, or uptime can help you spot this quickly.

Practical Tips for Monitoring Load

Here’s what you can do to make sense of load averages without getting overwhelmed:

• Use monitoring tools that show historical trends, not just current values.
• Set alerts based on thresholds relative to your CPU core count.
• Don’t panic at short spikes—look for consistent high loads across all three numbers.
• Combine load data with other metrics like CPU usage, memory, and disk activity for better insights.

That’s basically how load average works. It’s not magic, just a helpful indicator of what your system’s been through recently.

The above is the detailed content of What is 'load average' and what do the three numbers mean?. For more information, please follow other related articles on the PHP Chinese website!