SuperAPI AI Directory

prometheus-skill

General & other

View repository

Updated 10-16 09:55

About this skill

Basic Rules

When the user asks questions like "My id:xxx name:gis what problem does the service have now?", "Does the xxx_gis service have any problems?" or "Help me check the current status of the service with projectId=xxx, projectName=gis" and so on, wanting to query service monitoring metrics, please follow the 7 steps below to troubleshoot, and summarize and analyze the current status of the service based on the results (note to replace application).

  1. Query the current CPU usage Construct the promql statement and run prometheus.py, for example:

    uv run prometheus.py 'sum(rate(container_cpu_user_seconds_total{image!="",application="xxx_gis"}[30s])) by(application) * 100'
    

    Example result: "Latest value queried: 32.3487092130255" This means about 32% of CPU is being used (the maximum is not 100%, but the percentage of requested cores; for example, if 3 cores are requested, the maximum is 300%).

  2. Query the current CPU request amount Construct the promql statement and run prometheus.py, for example:

    uv run prometheus.py 'sum((container_spec_cpu_quota{application="xxx_gis", image!=""}) /1000) by (application)'
    

    Example result: "Latest value queried: 6181" This means this service has requested a total of 6181% CPU.

  1. Query the current Memory usage Construct the promql statement and run prometheus.py, for example:

    uv run prometheus.py 'sum(container_memory_rss{image!="",application="xxx_gis"}) by (application)'
    

    Example result: "Latest value queried: 21611335680.0" (unit is byte)

  2. Query the current Memory request amount Construct the promql statement and run prometheus.py, for example:

    uv run prometheus.py 'sum(container_spec_memory_limit_bytes{image!="",application="xxx_gis"}) by (application)'
    

    Example result: "Latest value queried: 75161927680.0" (unit is byte)

  3. Query the current Load Construct the promql statement and run prometheus.py, for example:

    uv run prometheus.py 'sum(container_cpu_load_average_10s{application="xxx_gis"}) by (application)'
    

    Example result: "Latest value queried: 1242.0", note that this value should be compared with the CPU request amount; in theory, less than 80% of the request amount is optimal. For example, if the CPU request amount is 2000, then 1242 is not high.

  4. Query the current HeapUsed usage Construct the promql statement and run prometheus.py, for example:

    uv run prometheus.py 'sum(jvm_memory_used_bytes{application="xxx_gis"}) by (application)'
    

    Example result: "Latest value queried: 15085025864.0" (unit byte)

  5. Query the current HeapMax maximum Construct the promql statement and run prometheus.py, for example:

    uv run prometheus.py 'sum(jvm_memory_max_bytes{application="xxx_gis"}) by (application)'
    

    Example result: "Latest value queried: 100503912448.0" (unit byte)