[BUG] CPU Metric of RDS not correct

tieungao88 commented 2 years ago

Is there an existing issue for this?

[X] I have searched the existing issues

Current Behavior

I used yet-another-cloudwatch-exporter for monitor RDS but metric yace (yet-another-cloudwatch-exporter) collect is not match with cloudwatch metric. Detail with CPU metric: In yace: 9.23% but in cloudwatch: 9.45%, period: 60s. This is my config:

  - type: rds
    enableMetricData: true
    regions:
      - ap-southeast-1
    roles:
      - roleArn: "arn:aws:iam::xxxxxxx:role/xxxx-yace-cloudwatch-role"
    metrics:
      - name: CPUUtilization
        statistics:
        - Average
        period: 60
        length: 60

Thanks.

Expected Behavior

No response

Steps To Reproduce

No response

Anything else?

No response

houlem commented 6 months ago

I had the same issue. Baffled for a long time one what the issue is. I added a random tag to the rds instance and restarted yace(was too impatient to wait for the next scrape interval) and it fixed it.

houlem commented 6 months ago

So the bug would be any untagged RDS resources are not scraped by Yace from Cloudwatch. Not sure if thats by design though.

prometheus-community / yet-another-cloudwatch-exporter