When the htop command is used, time + show red, is something wrong?

Question

below is my server htop display, the nginx process uses the processor time for more than 18 hours and shows red, but the CPU and memory look fine, is this value in the normal range?

Answer 1

I was also interested, so I dig into the source code and found this:

if (hours >= 100) { snprintf(buffer, 10, "%7lluh ", hours); RichString_append(str, CRT_colors[LARGE_NUMBER], buffer); } else { if (hours) { snprintf(buffer, 10, "%2lluh", hours); RichString_append(str, CRT_colors[LARGE_NUMBER], buffer); snprintf(buffer, 10, "%02d:%02d ", minutes, seconds); } else { snprintf(buffer, 10, "%2d:%02d.%02d ", minutes, seconds, hundredths); } RichString_append(str, CRT_colors[DEFAULT_COLOR], buffer); } 

Thus, it looks like as soon as the processor time exceeds one hour, part of the hour is simply highlighted in red (or something like CRT_colors[LARGE_NUMBER] ).

Note that the time format changes over time:

4:33.42 - minutes / seconds / millions

18h26:41 - hours / minutes / seconds

101h hours> 100

Answer 2

Most likely, just a notification method that will help you identify processes that show high CPU utilization within average load values. Check out the manual page.