fix: conflicting types of log fields in json output

[msvticket]

Summary

ScaledJob is an object in logs during setup, but a string later. So I change name to scaledJob.Name in the latter cases.

Details

When enabling json log format and debug level on the keda operator field ScaledObject has a value of type object during setup, but later is contains a string.

Examples

I have pretty printed the log lines for readability.

During setup

{
  "level": "info",
  "ts": "2024-03-12T05:00:08Z",
  "msg": "Reconciling ScaledJob",
  "controller": "scaledjob",
  "controllerGroup": "keda.sh",
  "controllerKind": "ScaledJob",
  "ScaledJob": {
    "name": "selenium-chrome-node",
    "namespace": "selenium-grid"
  },
  "namespace": "selenium-grid",
  "name": "selenium-chrome-node",
  "reconcileID": "a7191b5d-8677-43f2-adc2-47b285a448d9"
}

Later

{
  "level": "debug",
  "ts": "2024-03-12T05:00:09Z",
  "logger": "scalemetrics",
  "msg": "Checking if ScaleJob Scalers are active",
  "ScaledJob": "selenium-firefox-node",
  "isActive": false,
  "maxValue": 0,
  "MultipleScalersCalculation": ""
}

Effect

For me this prevents logs from being ingested into OpenSearch (using fluent bit), but it ought to cause problems with other objects stores as well.

Error message

Since I haven't explicitly set the type of the field ScaledObject in OpenSearch it will get the type object when the setup rows are ingested. Later this will cause errors like:

{
"type":"mapper_parsing_exception",
"reason":"object mapping for [ScaledJob] tried to parse field [ScaledJob] as object, but found a concrete value"
}

Checklist

• I have verified that my change is according to the deprecations & breaking changes policy
• Changelog has been updated and is aligned with our changelog requirements
• Commits are signed with Developer Certificate of Origin (DCO - learn more)

[zroubalik]

LGTM, thanks!

[zroubalik]

/run-e2e sj
Update: You can check the progress here

[kumarankit999]

Looks good that it resolves the Conflict types!!