fix: [sc-106256] Add missing uri field to troubleshoot.sh types

[nvanthao]

Story details: https://app.shortcut.com/replicated/story/106256
Demo: https://asciinema.org/a/c2tGPQrGZt6CM3jJQDZqKBsfv

This PR let preflight CLI load spec that has uri field, continued from #1574

I used Go reflect to extract all unique .Spec.uri in all troubleshoot objects. Once the PR is merged, we can perhaps update current logic in supportbundle CLI

troubleshoot/cmd/troubleshoot/cli/run.go

Lines 295 to 302 in 191ebdb

	// Load additional specs from support bundle URIs
	// only when no-uri flag is not set and no URLs are provided in the args
	if !viper.GetBool("no-uri") {
	err := loadSupportBundleSpecsFromURIs(ctx, kinds)
	if err != nil {
	klog.Warningf("unable to load support bundles from URIs: %v", err)
	}
	}

[banjoh]

When I tested this out, I could not replace the entire spec with what got downloaded from a uri. Instead, the specs get merged together

Original spec

apiVersion: troubleshoot.sh/v1beta2
kind: HostPreflight
metadata:
  name: my-spec
spec:
  uri: https://gist.githubusercontent.com/banjoh/142fabe108e199d42d442cc0017c7064/raw/a75ceff7f1ec40b58fc38c0dde70a20e9453b7f0/filesystem-preflight.yaml
  collectors:
  - diskUsage:
      collectorName: var-lib-kubelet
      path: /var/lib/kubelet

Wrongly updated spec

apiVersion: troubleshoot.sh/v1beta2
kind: HostPreflight
metadata:
  creationTimestamp: null
  name: my-spec
spec:
  analyzers:
  - filesystemPerformance:
      collectorName: Filesystem latency check
      outcomes:
      - pass:
          message: 'Write latency is ok (p99 target < 10ms, actual: {{ .P99 }})'
          when: p99 < 10ms
      - fail:
          message: Write latency is high. p99 target < 10ms, actual:{{ .String }}
  collectors:
  - diskUsage:
      collectorName: var-lib-kubelet
      path: /var/lib/kubelet
  - filesystemPerformance:
      backgroundIOPSWarmupSeconds: 10
      backgroundReadIOPS: 50
      backgroundReadIOPSJobs: 1
      backgroundWriteIOPS: 300
      backgroundWriteIOPSJobs: 6
      collectorName: Filesystem Latency Two Minute Benchmark
      datasync: true
      directory: /var/lib/etcd
      enableBackgroundIOPS: true
      fileSize: 22Mi
      timeout: 2m
  uri: https://gist.githubusercontent.com/banjoh/142fabe108e199d42d442cc0017c7064/raw/a75ceff7f1ec40b58fc38c0dde70a20e9453b7f0/filesystem-preflight.yaml
status: {}

Expected spec update

apiVersion: troubleshoot.sh/v1beta2
kind: HostPreflight
metadata:
  name: fs-preflight
spec:
  collectors:
    - filesystemPerformance:
        collectorName: Filesystem Latency Two Minute Benchmark
        timeout: 2m
        directory: /var/lib/etcd
        fileSize: 22Mi
        operationSizeBytes: 2300
        datasync: true
        enableBackgroundIOPS: true
        backgroundIOPSWarmupSeconds: 10
        backgroundWriteIOPS: 300
        backgroundWriteIOPSJobs: 6
        backgroundReadIOPS: 50
        backgroundReadIOPSJobs: 1
  analyzers:
    - filesystemPerformance:
        collectorName: Filesystem latency check
        outcomes:
          - pass:
              when: "p99 < 10ms"
              message: "Write latency is ok (p99 target < 10ms, actual: {{ .P99 }})"
          - fail:
              message: "Write latency is high. p99 target < 10ms, actual:{{ .String }}"

[nvanthao]

Hi @banjoh,

I've made changes to the code to follow our design doc. That is, to replace the original one entirely, not merge

Updated demo: https://asciinema.org/a/2n6nrYBIByC6SIuTsTSInX1sC

Please let me know how it goes.