SQS metricset reports misleading error message

[kaiyan-sheng]

User reports when running sqs metricset with config that specifies a endpoint, error message is still showing endpoint with amazonaws:

- module: aws
  period: 5m
  metricsets:
    - sqs
  #endpoint: "c2s.domain.tld"
  endpoint: "google.com"
  #endpoint: "askdlfjalsk.com"
  regions: "us-iso-east-1"
  access_key_id: "kasfdlkjasdflkasdjflkasdjf"

error message:

2020-12-03T15:20:58.405-0500    ERROR   [aws.sqs]       sqs/sqs.go:85   Error ListQueues: RequestError: send request failed
caused by: Post https://sqs.us-iso-east-1.amazonaws.com/: dial tcp: lookup sqs.us-iso-east-1.amazonaws.com on 192.168.1.1:53: no such host

[elasticmachine]

Pinging @elastic/integrations-platforms (Team:Platforms)

[bczika]

This issue was created on my behalf and a little more detail is needed as the description is not quite accurate. The problem is is the endpoint parameter is ALWAYS ignored and "amazonaws.com" is ALWAYS appended to the end of the url. I tried seven more aws modules in addition to sqs and all have the same problem therefore I'm certain the problem is not module specific.

[kaiyan-sheng]

This issue will be fixed with #23539. I tested with the PR and here is the new error message:

2021-01-18T12:35:52.520-0700    ERROR   [aws.sqs]       sqs/sqs.go:86   Error ListQueues: RequestError: send request failed
caused by: Post "https://sqs.us-iso-east-1.google.com/": dial tcp: lookup sqs.us-iso-east-1.google.com: no such host

[kaiyan-sheng]

This is fixed.