Prometheus average CPU utilization

Monitoring and Observability Prometheus
#1

Hi I am using testing the prometheus in aws. while i am testing the CPU usage. prometheus CPU usage values in not eqal with CLOud watch value.
i am using this query 100 - (avg by (instance) (irate(node_cpu_seconds_total{job="node",mode="idle"}[5m])) * 100) can you please anybody faced this issues. please help me/ I want get full(all cores of CPU) cpu usage

#2

maybe try […] avg by (instance, cpu) […]

#3

@UDBJM0RRS Thanks. I will try and let you know. Thanks

#4

You’re welcome. But be aware that the values may not completely reconcile due to sampling - but it should be close at least.

#5

Yup. I need closest value. Not exactly :slightly_smiling_face:

#6

Thanks @UDBJM0RRS. i got the output what i expected. Thanks lot :pray: