Skip to content

OCPEDGE-1036: fix: latency tuning for the rt-kernel tests on AWS metal #30790

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
jeff-roche wants to merge 1 commit into
base: main
Choose a base branch
from
+388 −70
Open

OCPEDGE-1036: fix: latency tuning for the rt-kernel tests on AWS metal #30790

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
43 changes: 39 additions & 4 deletions test/extended/kernel/common.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,9 +2,12 @@ package kernel

import (
"context"
"encoding/json"
"fmt"
"os"
"path/filepath"
"regexp"
"strings"
"time"

g "github.com/onsi/ginkgo/v2"
Expand All @@ -14,6 +17,7 @@ import (
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/labels"
"k8s.io/apimachinery/pkg/util/wait"
e2e "k8s.io/kubernetes/test/e2e/framework"
)

var (
Expand All @@ -37,17 +41,35 @@ func failIfNotRT(oc *exutil.CLI) {
o.Expect(len(rtNodes)).NotTo(o.BeZero(), "no realtime nodes are configured")
}

func getRealTimeWorkerNodes(oc *exutil.CLI) (nodes map[string]int, err error) {
nodes = make(map[string]int)

func getRealTimeWorkerNodes(oc *exutil.CLI) (nodes []string, err error) {
kubeNodes, err := oc.AsAdmin().KubeClient().CoreV1().Nodes().List(context.TODO(), metav1.ListOptions{LabelSelector: "node-role.kubernetes.io/worker="})
if err != nil {
return nodes, err
}

nodes = make([]string, 0, len(kubeNodes.Items))

nodesAreMetal := true

for _, node := range kubeNodes.Items {
if realTimeKernelRE.MatchString(node.Status.NodeInfo.KernelVersion) {
nodes[node.Name] = 0
nodes = append(nodes, node.Name)

nodeLabels := node.GetLabels()
if !strings.Contains(nodeLabels["node.kubernetes.io/instance-type"], "metal") {
nodesAreMetal = false
}
}
}

// Pad the latencies for non-metal instances
if !nodesAreMetal {
e2e.Logf("One or more nodes are not a metal instance, setting all real-time test thresholds to 7500 usec")
for test := range rtTestThresholds {
rtTestThresholds[test] = rtThresholdConfig{
SoftThreshold: 7500,
HardThreshold: 10000,
}
}
}

Expand Down Expand Up @@ -101,6 +123,19 @@ func cleanupRtTestPod(oc *exutil.CLI) {
o.Expect(err).NotTo(o.HaveOccurred(), "test pod for rt-tests never became ready")
}

// Write out a structured latency analysis artifact
func writeAnalysisArtifact(testName string, analysis *rtLatencyAnalysis) {
analysisJSON, err := json.MarshalIndent(analysis, "", " ")
if err != nil {
e2e.Logf("WARNING: unable to marshal analysis for %s: %v", testName, err)
return
}
writeTestArtifacts(
fmt.Sprintf("%s_%s_analysis.json", testName, e2e.TimeNow().Format(time.RFC3339)),
string(analysisJSON),
)
}

// Write out test artifacts
func writeTestArtifacts(fname string, content string) {
// Create the artifact dir for rt-tests if it does not exist
Expand Down
Loading