Confusing examples differences

[kaiohenricunha]

Some examples have this format, which is more in line with the sloth CRD:

apiVersion: sloth.slok.dev/v1
kind: PrometheusServiceLevel
metadata:
  name: sloth-slo-my-service
  namespace: monitoring
spec:
  service: "myservice"
  labels:
    owner: "myteam"
    repo: "myorg/myservice"
    tier: "2"
  slos:
    - name: "requests-availability"
      objective: 99.9
      description: "Common SLO based on availability for HTTP request responses."
      sli:
        events:
          errorQuery: sum(rate(http_request_duration_seconds_count{job="myservice",code=~"(5..|429)"}[{{.window}}]))
          totalQuery: sum(rate(http_request_duration_seconds_count{job="myservice"}[{{.window}}]))
      alerting:
        name: MyServiceHighErrorRate
        labels:
          category: "availability"
        annotations:
          summary: "High error rate on 'myservice' requests responses"
        pageAlert:
          labels:
            severity: pageteam
            routing_key: myteam
        ticketAlert:
          labels:
            severity: "slack"
            slack_channel: "#alerts-myteam"

Some others have this format:

version: "prometheus/v1"
service: "k8s-apiserver"
labels:
  cluster: "valhalla"
  component: "kubernetes"
slos:
  - name: "requests-availability"
    objective: 99.9
    description: "Warn that we are returning correctly the requests to the clients (kubectl users, controllers...)."
    labels:
      category: availability
    sli:
      events:
        error_query: sum(rate(apiserver_request_total{code=~"(5..|429)"}[{{.window}}]))
        total_query: sum(rate(apiserver_request_total[{{.window}}]))
    alerting:
      name: K8sApiserverAvailabilityAlert
      labels:
        category: "availability"
      annotations:
        runbook: "https://github.com/kubernetes-monitoring/kubernetes-mixin/tree/master/runbook.md#alert-name-kubeapierrorshigh"
      page_alert:
        labels:
          severity: critical
      ticket_alert:
        labels:
          severity: warning

  - name: "requests-latency"
    objective: 99
    description: "Warn that we apiserver responses are being slow and this will affect the clients  (kubectl users, controllers...)."
    labels:
      category: latency
    sli:
      events:
        error_query: |
          (
            sum(rate(apiserver_request_duration_seconds_count{verb!="WATCH"}[{{.window}}]))
            -
            sum(rate(apiserver_request_duration_seconds_bucket{le="0.4",verb!="WATCH"}[{{.window}}]))
          )
        total_query: sum(rate(apiserver_request_duration_seconds_count{verb!="WATCH"}[{{.window}}]))
    alerting:
      name: K8sApiserverLatencyAlert
      labels:
        category: "latency"
      annotations:
        runbook: "https://github.com/kubernetes-monitoring/kubernetes-mixin/tree/master/runbook.md#alert-name-kubeapilatencyhigh"
      page_alert:
        labels:
          severity: critical
      ticket_alert:
        labels:
          severity: warning

What is the correct way? What is this "version: "prometheus/v1""?

[markus-mnm]

sloth can be used from the command line or as a kubernetes operator.

From what I know, the cli is able to read all three specs as described here:
https://sloth.dev/specs/default/
Your first example, the Kubernetes spec (apiVersion: sloth.slok.dev/v1), is only usable by the sloth Kubernetes operator.
The second example (version: "prometheus/v1") is the default as used by the CLI