yet-another-cloudwatch-exporter: [Question] API GW V2 metrics not found with dimensionNameRequirements

Hello,

I have a question regarding the scraping of API GW V2 metrics. I noticed that the supporting PR was released in v0.52.0 for scraping API GW V2 metrics (HTTP API).

We have this config block in our config.yaml:

      # API GW V1 (REST)
      # https://docs.aws.amazon.com/apigateway/latest/developerguide/api-gateway-metrics-and-dimensions.html
      - <<: *default_job_config
        type: AWS/ApiGateway
        dimensionNameRequirements:
          - ApiName
          - Stage
        metrics:
          - name: Latency
            statistics: [ Average, Maximum, p95, p99 ]
          - name: Count
            statistics: [ SampleCount, Sum ]
          - name: 4XXError
            statistics: [ Sum, Average ]
          - name: 5XXError
            statistics: [ Sum, Average ]

      # API GW V2 (HTTP)
      # https://docs.aws.amazon.com/apigateway/latest/developerguide/http-api-metrics.html
      - <<: *default_job_config
        type: AWS/ApiGateway
        dimensionNameRequirements:
          - ApiId
          - Stage
          - Route
        metrics:
          - name: Latency
            statistics: [ Average, Maximum, p95, p99 ]
          - name: Count
            statistics: [ SampleCount, Sum ]
          - name: 4xx
            statistics: [ Sum ]
          - name: 5xx
            statistics: [ Sum ]

default_job_config has the values:

default_job_config: &default_job_config
  addCloudwatchTimestamp: true
  roles:
    - <roles>
  regions:
    - eu-west-1
  period: 300
  length: 300

We are also using these feature flags:

-enable-feature=aws-sdk-v2,max-dimensions-associator,list-metrics-callback

But we are getting these logs when running (multiple times):

{"account":"<account>","arn":"<role>","job_type":"AWS/ApiGateway","level":"info","msg":"No metrics data found","region":"eu-west-1","time":"2023-06-22T13:09:42Z"}

What are we doing wrong in our config.yaml?

About this issue

Original URL
State: open
Created a year ago
Comments: 29 (2 by maintainers)

Most upvoted comments

I’ve dumped a potential fix for the feature flag maxdimassociator in https://github.com/nerdswords/yet-another-cloudwatch-exporter/pull/1081

It seems to work correctly now, but you need https://github.com/nerdswords/yet-another-cloudwatch-exporter/pull/1067 as well. I’ll try to get both merged tomorrow and so you can test a build from master.

cristiangreco on Jul 31, 2023

Hello @cristiangreco,

I just tested #1067 ~~and it seems to work! In the debug output, the labels: Method, Stage and Resource are being registered and shown correctly!~~

Sorry, I misinterpret the debug output, after testing #1067, it seems that it still does not work. From the /metrics page, I still see that the labels method, stage and resource are missing:

aws_apigateway_latency_average{account_id="xxx",dimension_ApiId="xxx",name="arn:aws:apigateway:eu-west-1::/apis/xxx",region="eu-west-1"}

I am seeing this in debug logging output:

{"account":"xxx","arn":"","dimensions":"Stage=$default,ApiId=xxx","job_type":"AWS/ApiGateway","level":"debug","metric":"5xx","msg":"skipping metric unmatched by associator","region":"eu-west-1","time":"2023-07-26T13:36:28+02:00"}
{"account":"xxx","arn":"","dimensions":"Resource=/{proxy+},Stage=$default,Method=OPTIONS,ApiId=xxx","job_type":"AWS/ApiGateway","level":"debug","metric":"5xx","msg":"skipping metric unmatched by associator","region":"eu-west-1","time":"2023-07-26T13:36:28+02:00"}

bogdankatishev on Jul 26, 2023