current total CPU time is less than previous total CPU time

[lifehacking]

Relevant config.toml

cpu.toml

# # collect interval
# interval = 15

# # whether collect per cpu
# collect_per_cpu = false

Logs from categraf

2025/01/08 10:22:47 cpu.go:73: W! current total CPU time is less than previous total CPU time
2025/01/08 10:22:55 heartbeat.go:219: W! current total CPU time is less than previous total CPU time
10:23:02 cpu_usage_iowait agent_hostname=192.168.1.6 cpu=cpu-total 6.671085684330087e-11
10:23:02 cpu_usage_idle agent_hostname=192.168.1.6 cpu=cpu-total 6.184096428312938e-07
10:23:02 cpu_usage_nice agent_hostname=192.168.1.6 cpu=cpu-total 0
10:23:02 cpu_usage_irq agent_hostname=192.168.1.6 cpu=cpu-total 0
10:23:02 cpu_usage_softirq agent_hostname=192.168.1.6 cpu=cpu-total 7.671748535377828e-09
10:23:02 cpu_usage_steal agent_hostname=192.168.1.6 cpu=cpu-total 99.99999923035683
10:23:02 cpu_usage_guest agent_hostname=192.168.1.6 cpu=cpu-total 0
10:23:02 cpu_usage_user agent_hostname=192.168.1.6 cpu=cpu-total 1.2288139827982894e-07
10:23:02 cpu_usage_system agent_hostname=192.168.1.6 cpu=cpu-total 2.061365476093958e-08
10:23:02 cpu_usage_guest_nice agent_hostname=192.168.1.6 cpu=cpu-total 0
10:23:02 cpu_usage_active agent_hostname=192.168.1.6 cpu=cpu-total 99.99999938159036
2025/01/08 10:23:05 heartbeat.go:219: W! current total CPU time is less than previous total CPU time
2025/01/08 10:23:15 heartbeat.go:219: W! current total CPU time is less than previous total CPU time

System info

kvm centos7/categraf-v0.3.37

Docker

No response

Steps to reproduce

when categraf startup ，some days ago，cpu_usage_active metrics exception，cpu_usage_active value is 99.99xxxx，it‘’s not true value for system os cpu useage

Expected behavior

cpu_usage_active metrics exception，it‘’s not system os cpu metric true value

Actual behavior

show true value

Additional info

No response

[UlricQin]

你用中文描述问题吧。看不懂。cpu_usage_active 这个指标是 CPU 使用率，有时大有时小是正常的，cpu_usage_active 并非是 OS 启动以来 CPU 使用的秒数，和 node_exporter 的指标不同。