e2e: consistent failure logging

All failures are worth logging immediately, not just unexpected
errors. That helps understand tests that have long-running cleanup
operations with their own logging, because the failure will be visible
inside the test output.

The logging in framework.ExpectNoError also was rather poor, because
it only showed the error, but not the additional information about it.

Tests suites now should use log.Fail as Gomega failure handler instead
of ginkgowrapper.Fail. log.Fail will handle the logging for all
failures before proceeding to record the failure in Ginkgo.

Because logging is always done also after a test failure, additional
failures during cleanup are now visible. Ginkgo itself just ignores
them.

test/e2e/e2e.go

@@ -32,7 +32,7 @@ import (
 	"k8s.io/component-base/logs"
 	commontest "k8s.io/kubernetes/test/e2e/common"
 	"k8s.io/kubernetes/test/e2e/framework"
-	"k8s.io/kubernetes/test/e2e/framework/ginkgowrapper"
+	e2elog "k8s.io/kubernetes/test/e2e/framework/log"
 
 	// ensure auth plugins are loaded
 	_ "k8s.io/client-go/plugin/pkg/client/auth"
@@ -71,7 +71,7 @@ func RunE2ETests(t *testing.T) {
 	logs.InitLogs()
 	defer logs.FlushLogs()
 
-	gomega.RegisterFailHandler(ginkgowrapper.Fail)
+	gomega.RegisterFailHandler(e2elog.Fail)
 	// Disable skipped tests unless they are explicitly requested.
 	if config.GinkgoConfig.FocusString == "" && config.GinkgoConfig.SkipString == "" {
 		config.GinkgoConfig.SkipString = `\[Flaky\]|\[Feature:.+\]`

test/e2e/framework/log/logger.go

@@ -48,6 +48,17 @@ func Failf(format string, args ...interface{}) {
 // (for example, for call chain f -> g -> FailfWithOffset(1, ...) error would be logged for "f").
 func FailfWithOffset(offset int, format string, args ...interface{}) {
 	msg := fmt.Sprintf(format, args...)
-	log("INFO", msg)
+	log("FAIL", msg)
 	ginkgowrapper.Fail(nowStamp()+": "+msg, 1+offset)
 }
+
+// Fail is a replacement for ginkgo.Fail which logs the problem as it occurs
+// and then calls ginkgowrapper.Fail.
+func Fail(msg string, callerSkip ...int) {
+	skip := 1
+	if len(callerSkip) > 0 {
+		skip += callerSkip[0]
+	}
+	log("FAIL", msg)
+	ginkgowrapper.Fail(nowStamp()+": "+msg, skip)
+}

test/e2e/framework/log/logger_test.go

@@ -29,7 +29,6 @@ import (
 	"github.com/onsi/gomega"
 
 	"k8s.io/kubernetes/test/e2e/framework"
-	"k8s.io/kubernetes/test/e2e/framework/ginkgowrapper"
 	"k8s.io/kubernetes/test/e2e/framework/log"
 )
 
@@ -46,7 +45,7 @@ var _ = ginkgo.Describe("log", func() {
 		gomega.Expect(false).To(gomega.Equal(true), "false is never true")
 	})
 	ginkgo.It("error", func() {
-		err := errors.New("I'm an error, nice to meet to.")
+		err := errors.New("an error with a long, useless description")
 		framework.ExpectNoError(err, "hard-coded error")
 	})
 	ginkgo.AfterEach(func() {
@@ -60,39 +59,50 @@ func TestFailureOutput(t *testing.T) {
 	// reporter in adddition to the default one. To see what the full
 	// Ginkgo report looks like, run this test with "go test -v".
 	config.DefaultReporterConfig.FullTrace = true
-	gomega.RegisterFailHandler(ginkgowrapper.Fail)
+	gomega.RegisterFailHandler(log.Fail)
 	fakeT := &testing.T{}
 	reporter := reporters.NewFakeReporter()
 	ginkgo.RunSpecsWithDefaultAndCustomReporters(fakeT, "Logging Suite", []ginkgo.Reporter{reporter})
 
 	// Now check the output.
+	// TODO: all of the stacks are currently broken because Ginkgo doesn't properly skip
+	// over the initial entries returned by runtime.Stack. Fix is pending in
+	// https://github.com/onsi/ginkgo/pull/590, "stack" texts need to be updated
+	// when updating to a fixed Ginkgo.
 	g := gomega.NewGomegaWithT(t)
-	g.Expect(normalizeReport(*reporter)).To(gomega.Equal(suiteResults{
+	actual := normalizeReport(*reporter)
+	expected := suiteResults{
 		testResult{
-			name: "[Top Level] log asserts",
-			// TODO: also log the failed assertion as it happens
-			output:  "INFO: before\nINFO: after\n",
+			name:    "[Top Level] log asserts",
+			output:  "INFO: before\nFAIL: false is never true\nExpected\n    <bool>: false\nto equal\n    <bool>: true\nINFO: after\nFAIL: true is never false either\nExpected\n    <bool>: true\nto equal\n    <bool>: false\n",
 			failure: "false is never true\nExpected\n    <bool>: false\nto equal\n    <bool>: true",
-			// TODO: ginkowrapper.Fail should also prune this stack.
-			stack: "\tassertion.go:75\nk8s.io/kubernetes/vendor/github.com/onsi/gomega/internal/assertion.(*Assertion).To()\n\tassertion.go:38\nk8s.io/kubernetes/test/e2e/framework/log_test.glob..func1.3()\n\tlogger_test.go:46\nk8s.io/kubernetes/vendor/github.com/onsi/ginkgo/internal/leafnodes.(*runner).runSync()\n\tlogger_test.go:66\n",
+			// TODO: should start with k8s.io/kubernetes/test/e2e/framework/log_test.glob..func1.3()
+			stack: "\tassertion.go:75\nk8s.io/kubernetes/vendor/github.com/onsi/gomega/internal/assertion.(*Assertion).To()\n\tassertion.go:38\nk8s.io/kubernetes/test/e2e/framework/log_test.glob..func1.3()\n\tlogger_test.go:45\nk8s.io/kubernetes/vendor/github.com/onsi/ginkgo/internal/leafnodes.(*runner).runSync()\n\tlogger_test.go:65\n",
 		},
 		testResult{
-			name: "[Top Level] log error",
-			// TODO: the additional information about the error should be logged
-			output:  "INFO: before\nINFO: Unexpected error occurred: I'm an error, nice to meet to.\nINFO: after\n",
-			failure: "hard-coded error\nUnexpected error:\n    <*errors.errorString>: {\n        s: \"I'm an error, nice to meet to.\",\n    }\n    I'm an error, nice to meet to.\noccurred",
-			// TODO: ginkowrapper.Fail should also prune this stack.
+			name:    "[Top Level] log error",
+			output:  "INFO: before\nFAIL: hard-coded error\nUnexpected error:\n    <*errors.errorString>: {\n        s: \"an error with a long, useless description\",\n    }\n    an error with a long, useless description\noccurred\nINFO: after\nFAIL: true is never false either\nExpected\n    <bool>: true\nto equal\n    <bool>: false\n",
+			failure: "hard-coded error\nUnexpected error:\n    <*errors.errorString>: {\n        s: \"an error with a long, useless description\",\n    }\n    an error with a long, useless description\noccurred",
+			// TODO: should start with k8s.io/kubernetes/test/e2e/framework/log_test.glob..func1.4()
 			stack: "\tutil.go:1362\nk8s.io/kubernetes/test/e2e/framework.ExpectNoError()\n\tutil.go:1356\nk8s.io/kubernetes/test/e2e/framework/log_test.glob..func1.4()\n\tlogger_test.go:49\nk8s.io/kubernetes/vendor/github.com/onsi/ginkgo/internal/leafnodes.(*runner).runSync()\n\tlogger_test.go:65\n",
 		},
 		testResult{
-			name: "[Top Level] log fails",
-			// TODO: why is the failure log as "INFO"?
-			output:  "INFO: before\nINFO: I'm failing.\nINFO: after\n",
+			name:    "[Top Level] log fails",
+			output:  "INFO: before\nFAIL: I'm failing.\nINFO: after\nFAIL: true is never false either\nExpected\n    <bool>: true\nto equal\n    <bool>: false\n",
 			failure: "I'm failing.",
-			// TODO: ginkowrapper.Fail should also prune this stack.
+			// TODO: should start with k8s.io/kubernetes/test/e2e/framework/log_test.glob..func1.2.1(...)
 			stack: "\tlogger.go:52\nk8s.io/kubernetes/test/e2e/framework/log.Failf()\n\tlogger.go:44\nk8s.io/kubernetes/test/e2e/framework/log_test.glob..func1.2.1(...)\n\tlogger_test.go:41\nk8s.io/kubernetes/test/e2e/framework/log_test.glob..func1.2()\n\tlogger_test.go:42\nk8s.io/kubernetes/vendor/github.com/onsi/ginkgo/internal/leafnodes.(*runner).runSync()\n\tlogger_test.go:65\n",
 		},
-	}))
+	}
+	// Compare individual fields. Comparing the slices leads to unreadable error output when there is any mismatch.
+	g.Expect(len(actual)).To(gomega.Equal(len(expected)), "%d entries in %v", len(expected), actual)
+	for i, a := range actual {
+		b := expected[i]
+		g.Expect(a.name).To(gomega.Equal(b.name), "name in %d", i)
+		g.Expect(a.output).To(gomega.Equal(b.output), "output in %d", i)
+		g.Expect(a.failure).To(gomega.Equal(b.failure), "failure in %d", i)
+		g.Expect(a.stack).To(gomega.Equal(b.stack), "stack in %d", i)
+	}
 }
 
 type testResult struct {

test/e2e/framework/util.go

@@ -1359,10 +1359,6 @@ func ExpectNoError(err error, explain ...interface{}) {
 // ExpectNoErrorWithOffset checks if "err" is set, and if so, fails assertion while logging the error at "offset" levels above its caller
 // (for example, for call chain f -> g -> ExpectNoErrorWithOffset(1, ...) error would be logged for "f").
 func ExpectNoErrorWithOffset(offset int, err error, explain ...interface{}) {
-	if err != nil {
-		e2elog.Logf("Unexpected error occurred: %v", err)
-		debug.PrintStack()
-	}
 	gomega.ExpectWithOffset(1+offset, err).NotTo(gomega.HaveOccurred(), explain...)
 }