feat: refactor minikube test bot message

[ComradeProgrammer]

After:

Here are the number of top 10 failed tests in each environments with lowest flake rate.

Environment	Test Name	Flake Rate
Docker_Linux_crio_arm64 (3 failed)	TestMultiControlPlane/serial/RestartCluster(gopogh)	11.56% (chart)

Besides the following environments also have failed tests:

• Docker_Windows: 3 failed (gopogh)

• Docker_Linux_containerd_arm64: 6 failed (gopogh)

• Hyper-V_Windows: 14 failed (gopogh)

• Docker_Linux_docker_arm64: 1 failed (gopogh)

• Docker_Linux_crio: 2 failed (gopogh)

• QEMU_macOS: 156 failed (gopogh)

• Docker_macOS: 1 failed (gopogh)

• KVM_Linux_crio: 31 failed (gopogh)

To see the flake rates of all tests by environment, click here.

[ComradeProgrammer]

/ok-to-test

[medyagh]

@ComradeProgrammer lets make it in a way that if it has more than 10 failures in an environment it should be skipped and instead just a english message that this enviroment had ALOT of failures...

[spowelljr]

Looking at your example output:

Environment	Test Name	Flake Rate
Docker_Linux_crio (2 failed)	TestAddons/parallel/Ingress(gopogh)	100.00% (chart)
Docker_Linux_crio (2 failed)	TestAddons/parallel/MetricsServer(gopogh)	100.00% (chart)
Docker_macOS (1 failed)	TestSkaffold(gopogh)	93.10% (chart)

I think we should have a flake threshold that we don't show. Showing flakes that are at or near 100% isn't very helpful, it's just noise that we have to read through to find failures that we care about. In the existing flake rate comment we only show tests that have less than 50% flake rate if ($3 < 50) printf "%s:%s,%s\n", $1, $2, $3, I think we should implement something similar.

[medyagh]

Looking at your example output:

Environment Test Name Flake Rate
Docker_Linux_crio (2 failed) TestAddons/parallel/Ingress(gopogh) 100.00% (chart)
Docker_Linux_crio (2 failed) TestAddons/parallel/MetricsServer(gopogh) 100.00% (chart)
Docker_macOS (1 failed) TestSkaffold(gopogh) 93.10% (chart)
I think we should have a flake threshold that we don't show. Showing flakes that are at or near 100% isn't very helpful, it's just noise that we have to read through to find failures that we care about. In the existing flake rate comment we only show tests that have less than 50% flake rate if ($3 < 50) printf "%s:%s,%s\n", $1, $2, $3, I think we should implement something similar.

thats a a good idea

[ComradeProgrammer]

Now tests with >50% flake rate will not be shown

[medyagh]

thank you

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: ComradeProgrammer, medyagh

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [medyagh]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[spowelljr]

Suggested change

	if len(records[i]) < 2 {
	if len(records[i]) < 3 {

records[i][2] is below, so need to increase length check

[medyagh]

@ComradeProgrammer can you plz take a look

[spowelljr]

Good work, looks really good

[spowelljr]

Suggested change

	client, err := storage.NewClient(context.Background())
	client, err := storage.NewClient(ctx)

[spowelljr]

Suggested change

	btk := client.Bucket("minikube-builds")
	obj := btk.Object(fmt.Sprintf("logs/%s/%s/%s_summary.json", pr, rootJob, env))
	bkt := client.Bucket("minikube-builds")
	obj := bkt.Object(fmt.Sprintf("logs/%s/%s/%s_summary.json", pr, rootJob, env))

[spowelljr]

Suggested change

	btk := client.Bucket("minikube-flake-rate")
	obj := btk.Object("flake_rates.csv")
	bkt := client.Bucket("minikube-flake-rate")
	obj := bkt.Object("flake_rates.csv")

[spowelljr]

Suggested change

	// the csv must have at least 2 columns
	// the csv must have at least 3 columns

[spowelljr]

Suggested change

	// flakeRate downloads recent flake rates from GCS, and returns a map{env->map{testname->flake rate}}
	// flakeRate downloads recent flake rates from GCS, and returns a
	// map{env->map{testname->flake rate}}

Shortening long comments for better readability

[minikube-pr-bot]

kvm2 driver with docker runtime

+----------------+----------+---------------------+
|    COMMAND     | MINIKUBE | MINIKUBE (PR 19036) |
+----------------+----------+---------------------+
| minikube start | 50.1s    | 50.2s               |
| enable ingress | 25.0s    | 25.7s               |
+----------------+----------+---------------------+

Times for minikube ingress: 23.9s 24.4s 27.9s 24.9s 23.9s
Times for minikube (PR 19036) ingress: 24.4s 24.9s 26.9s 27.4s 24.9s

Times for minikube start: 47.7s 51.1s 51.8s 50.2s 49.8s
Times for minikube (PR 19036) start: 48.5s 49.1s 50.2s 49.6s 53.4s

docker driver with docker runtime

+----------------+----------+---------------------+
|    COMMAND     | MINIKUBE | MINIKUBE (PR 19036) |
+----------------+----------+---------------------+
| minikube start | 21.3s    | 22.1s               |
| enable ingress | 21.5s    | 22.0s               |
+----------------+----------+---------------------+

Times for minikube start: 21.3s 21.3s 21.0s 21.6s 21.5s
Times for minikube (PR 19036) start: 20.8s 23.0s 23.5s 19.7s 23.5s

Times for minikube (PR 19036) ingress: 21.7s 22.8s 21.8s 21.8s 21.8s
Times for minikube ingress: 21.8s 21.2s 21.3s 21.8s 21.2s

docker driver with containerd runtime

+----------------+----------+---------------------+
|    COMMAND     | MINIKUBE | MINIKUBE (PR 19036) |
+----------------+----------+---------------------+
| minikube start | 21.2s    | 20.9s               |
| enable ingress | 29.8s    | 32.1s               |
+----------------+----------+---------------------+

Times for minikube start: 23.5s 19.6s 23.0s 19.8s 19.9s
Times for minikube (PR 19036) start: 19.8s 20.5s 22.2s 22.7s 19.4s

Times for minikube (PR 19036) ingress: 32.2s 31.8s 32.2s 32.2s 32.2s
Times for minikube ingress: 33.3s 32.3s 31.7s 32.8s 18.8s