When you run htop on a Linux system, you'll notice three distinct load average values displayed near the top of the interface. These represent:

Load average: 1.24 0.89 0.67

The three numbers show the system load over:

1. The last 1 minute (bold)

2. The last 5 minutes (normal)

3. The last 15 minutes (red)

The colors indicate urgency levels:

Bold (1-minute): Immediate system state

Normal (5-minute): Recent trend

Red (15-minute): Baseline for comparison

Consider this output from a server:

Load average: 4.32 1.85 0.92

This shows:

- Current spike (4.32)

- Gradual increase from 15m (0.92) to 5m (1.85)

- Possible short-lived process affecting performance

Here's a Python script to extract and analyze these values:

import os

def get_load_averages():
    with open('/proc/loadavg') as f:
        load = f.read().split()
    return {
        '1min': float(load[0]),
        '5min': float(load[1]),
        '15min': float(load[2])
    }

loads = get_load_averages()
print(f"Current load: {loads['1min']:.2f}")
if loads['1min'] > loads['15min'] * 2:
    print("Warning: Recent load spike detected!")

Compare load averages to your CPU cores:

- 4-core system with load of 4: 100% utilization

- Same system with load of 8: processes queuing

- Consistently high 15-minute average indicates chronic issues

You can modify the colors in ~/.config/htop/htoprc:

# Change load average colors
color_scheme=0
load1_color=6
load5_color=2
load15_color=1

When you run htop (or check traditional uptime), you'll notice three numbers representing system load averages. These represent the system load over the last 1-minute, 5-minute, and 15-minute intervals respectively. The color coding in htop serves as visual indicators:

1. Bold (usually white): 1-minute average
2. Normal: 5-minute average  
3. Red: 15-minute average

Contrary to common belief, load averages don't represent CPU usage percentage. They indicate the number of processes:

• Currently running on CPU cores
• Waiting for CPU time (runnable but not executing)
• Blocked on uninterruptible I/O operations

For example, on a 4-core system:

Load average: 4.00 means fully utilized
Load average: 8.00 means processes waiting (2x overload)
Load average: 2.00 means 50% idle capacity

htop uses color to indicate temporal trends:

Red coloring typically triggers when the 15-minute average exceeds:

Number of CPU cores × load threshold (usually 0.7)

Here's how to check load averages programmatically in Python:

import os

def get_load_averages():
    with open('/proc/loadavg') as f:
        loads = f.read().split()[:3]
    return [float(x) for x in loads]

one_min, five_min, fifteen_min = get_load_averages()

In Bash, you can extract specific values:

# Get 15-minute average
cat /proc/loadavg | awk '{print $3}'

# Alternative using uptime
uptime | grep -oP '(?<=load average: ).*'