Source code paths during //test/e2e/framework/log:go_default_test in
the Kubernetes CI start with relative paths. To avoid too broad
matching of the regex, those paths that occur in practice are named
explicitly as alternatives to the leading slash.
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.
This test runs a fake Ginkgo suite with various errors and checks how
logger.go respectively ginkgowrapper.go handle this. Right now, the
outcome is sub-optimal:
- some test failures (those that use framework.Failf or
framework.ExpectNoError) are logged immediately, Gomega failures
are not
- framework.ExpectNoError logs just the error, which is often useless
without the additional explanation
- failures that occur after some others are not reported at all;
this can happen in cleanup code and while that code should better
be written so that it contines instead of failing on an assertion,
in practice quite a lot of code fails and when it does, the output
would be useful to have
- the full stack trace is odd and doesn't start with the expected
function name
Example output:
• Failure [0.002 seconds]
log
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:35
fails [It]
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:39
Jul 17 12:00:52.545: I'm failing.
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:41
Full Stack Trace
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger.go:51 +0x143
k8s.io/kubernetes/test/e2e/framework/log.Failf(...)
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger.go:43
k8s.io/kubernetes/test/e2e/framework/log_test.glob..func1.2.1(...)
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:41
k8s.io/kubernetes/test/e2e/framework/log_test.glob..func1.2()
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:42 +0x52
k8s.io/kubernetes/vendor/github.com/onsi/ginkgo/internal/leafnodes.(*runner).runSync(0xc00029b020, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, ...)
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:65 +0x1c8
testing.tRunner(0xc000358600, 0x19818c0)
/nvme/gopath/go/src/testing/testing.go:865 +0xc0
created by testing.(*T).Run
/nvme/gopath/go/src/testing/testing.go:916 +0x35a
------------------------------
Jul 17 12:00:52.545: INFO: before
Jul 17 12:00:52.545: INFO: I'm failing.
Jul 17 12:00:52.547: INFO: after
• Failure [0.002 seconds]
log
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:35
asserts [It]
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:44
false is never true
Expected
<bool>: false
to equal
<bool>: true
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:45
Full Stack Trace
/nvme/gopath/src/k8s.io/kubernetes/vendor/github.com/onsi/gomega/internal/assertion/assertion.go:75 +0x1f1
k8s.io/kubernetes/vendor/github.com/onsi/gomega/internal/assertion.(*Assertion).To(0xc00035f6c0, 0x1b42140, 0xc000350dd0, 0xc000350de0, 0x1, 0x1, 0x42e35f)
/nvme/gopath/src/k8s.io/kubernetes/vendor/github.com/onsi/gomega/internal/assertion/assertion.go:38 +0xc7
k8s.io/kubernetes/test/e2e/framework/log_test.glob..func1.3()
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:45 +0x17e
k8s.io/kubernetes/vendor/github.com/onsi/ginkgo/internal/leafnodes.(*runner).runSync(0xc00029b0e0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, ...)
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:65 +0x1c8
testing.tRunner(0xc000358600, 0x19818c0)
/nvme/gopath/go/src/testing/testing.go:865 +0xc0
created by testing.(*T).Run
/nvme/gopath/go/src/testing/testing.go:916 +0x35a
------------------------------
Jul 17 12:00:52.548: INFO: before
Jul 17 12:00:52.549: INFO: after
• Failure [0.002 seconds]
log
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:35
error [It]
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:47
hard-coded error
Unexpected error:
<*errors.errorString | 0xc000351930>: {
s: "I'm an error, nice to meet to.",
}
I'm an error, nice to meet to.
occurred
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:49
Full Stack Trace
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/util.go:1376 +0x191
k8s.io/kubernetes/test/e2e/framework.ExpectNoError(...)
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/util.go:1367
k8s.io/kubernetes/test/e2e/framework/log_test.glob..func1.4()
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:49 +0xc9
k8s.io/kubernetes/vendor/github.com/onsi/ginkgo/internal/leafnodes.(*runner).runSync(0xc00029b200, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, ...)
/nvme/gopath/src/k8s.io/kubernetes/test/e2e/framework/log/logger_test.go:65 +0x1c8
testing.tRunner(0xc000358600, 0x19818c0)
/nvme/gopath/go/src/testing/testing.go:865 +0xc0
created by testing.(*T).Run
/nvme/gopath/go/src/testing/testing.go:916 +0x35a
------------------------------
Jul 17 12:00:52.550: INFO: before
Jul 17 12:00:52.550: INFO: Unexpected error occurred: I'm an error, nice to meet to.
Jul 17 12:00:52.551: INFO: after
This makes sub packages of e2e test framework to use log functions
of core framework instead for avoiding circular dependencies.
NOTE: subpackage pod would make circular dependency if changing here.
So we need to take care of it with another PR.
PrintPerfData is called at e2e node tests and it depends on e2elog
and e2emetrics. This moves the function to the place which calls
the function for removing unnecessary dependencies from e2e node
subpackage.
Currently, Kubernetes supports running as different user (RunAsUser),
but it only supports UIDs, which does not work on Windows.
Which is why the field SecurityContext.WindowsOptions.RunAsUserName
was introduced, to allow us to run the container entrypoints with
a different user than its default one.
This commit adds E2E tests which will validate this behaviour. The
tests are Windows only, and they will be skipped if --node-os-distro
is not "windows".
This makes sub packages of e2e test framework to use log functions
of core framework instead for avoiding circular dependencies.
NOTE: test/e2e/framework/kubelet, test/e2e/framework/metrics and
test/e2e/framework/node will make circular dependencies if
updating them. It is necessary to solve them in advance before
this work.
Promotes the VolumePVCDataSource feature (cloning) to beta for the 1.16
release.
Since alpha release in 1.15 there have been a number of minor bug fixes
in the CSI Hospath Provisioner and the CSI provisioner sidecar. We've
also added e2e tests using the Hostpath provisioner.
If requesting an substantial amount of huge page memory
(like the tests), it is recommended to to it as early
after boot as possible, or by using the kernel cmd line.
There are some functions of e2e test framework and it is useful to
read the test code by using these functions.
This replaces gomega calls with these functions under test/e2e/node/
Skips IPv6 tests on Windows.
Skips sysctl tests on Windows.
Skips network policy tests on Windows.
Skips RunAsUser / FSGroup / file permissions related tests, as those are
not supported on Windows.
Skips the test "should preserve source pod IP for traffic thru service cluster IP"
on Windows, as it creates a Pod with HostNetwork=true, which is unsupported.
What works and what doesn't work on Windows has been documented here:
https://github.com/kubernetes/community/blob/master/contributors/devel/sig-architecture/conformance-tests.md#windows--linux-considerations
Add a Feature:HPA tag to these tests so they're not picked up by
the release-blocking job that focuses on [Serial] tests (but
excludes [Feature:.*] tests)
They take a combined 70 minutes on average. If they really need
to be in release-blocking as implemented, we should consider a
separate job to focus just on this feature.
Add a Feature:RegularResourceUsageTracking tag to these tests so
they're not picked up by the release-blocking job that focuses
on [Serial] tests (but excludes [Feature:.*] tests)
They take a combined 65 minutes on average. If they really need
to be in release-blocking as implemented, we should consider a
separate job to focus just on this feature.
HandleFlags() was used at e2e package and it depends on sub e2e
framework "config" in core e2e framework. That was invalid dependency.
So this moves HandleFlags() to e2e package for simple dependency.
We tried to separate logger functionality as subpackage of e2e test
framework, but we've recognized that should be core functionality
and we should keep it as core of e2e test framework after facing
circular dependency issues.
So this adds log.go back to core of e2e test framework. In addition,
this makes volume sub package use the core logger as a sample.
Also remove FS VolMode disruptive tests because they are already covered
through the subpath disruptive test cases.
Change-Id: Ice4b30b0d8fdcb1f7fd61e54d27f53557de9f13a
ConfirmStatefulPodCount() was used at e2e statefulset test only,
and that added dependency on another sub framework to statefulset
sub framework. This moves ConfirmStatefulPodCount() to the e2e test
for clean dependency.
Includes the changes from #80922 (that were reverted), and updates the test to appropriately add intolerable taints to all nodes except the target node
DesiredStateOfWorldPopulator should skip a volume that is not used in any
pod. "Used" means either mounted (via volumeMounts) or used as raw block
device (via volumeDevices).
Especially when block feature is disabled, a block volume must not get into
DesiredStateOfWorld, because it would be formatted and mounted there.
Every caller of ReadOrDie() specified ginkgo.Fail as fail argument,
and that was intentional to avoid depending on Ginkgo.
However that just spreaded the dependency on Ginkgo to caller sides.
Especially that was unnecessary e2e test framework "ingress" depended
on Ginkgo only for the above reason.
Now we are cleaning up the dependencies on e2e tests, so let's just
remove such dependencies.
GetGPUDevicePluginImage() was called in some e2e node test only.
So it is not worth keeping the function as a part of e2e test
framework. This moves GetGPUDevicePluginImage() to the e2e node
test for code cleanup.
In e2e test framework, ProxyRequest() is called in kubelet e2e test
framework only. In addition, the function works for kubelet as tests
which call the function do. So this moves the function into kubelet
e2e test framework to make the dependency simple.
Using a "normal" CSI driver for an inline ephemeral volume may have
unexpected and potentially harmful effects when the driver gets a
NodePublishVolume call that it isn't expecting. To prevent that mistake,
driver deployments for a driver that supports such volumes must:
- deploy a CSIDriver object for the driver
- set CSIDriver.Spec.VolumeLifecycleModes such that it contains "ephemeral"
The default for that field is "persistent", so existing deployments
continue to work and are automatically protected against incorrect
usage.
For the E2E tests we need a way to specify the driver mode. The
existing cluster-driver-registrar doesn't support that and also was
deprecated, so we stop using it altogether and instead deploy and
patch a CSIDriver object.
A previous commit replaced the usage of hostname -i with getent hosts
in DNS probe commands (for testing IPv6), which breaks the Windows tests
(getent hosts is a Linux-specific command that looks into the /etc/hosts
file, which, on Windows, is not managed by Kubelet).
This commit addresses this issue by executing getent hosts on IPv6 clusters
and hostname -i on IPv4 clusters (Windows does not support IPv6 at this
moment).
The core e2e framework contains BusyBoxImage. The sub e2e "pod"
framework also contains it but not used at all.
So this removes this unused BusyBoxImage for code cleanup.
WaitForEndpoint() of the endpoints e2e framework was used in
test/e2e/network/proxy.go only. In addition, the endpoints e2e
framework imported the core of e2e framework only for the function.
So this moves the function into test/e2e/network/proxy.go then we
can remove dependency of core framework from the sub e2e framework.
RunDeployment() of the deployment e2e framework was used in
autoscaling_utils only. In addition, the deployment e2e framework
imported the core of e2e framework only for the function.
So this moves the function into autoscaling_utils then we can remove
dependency of core framework from the deployment e2e framework.
all of these folks are either approvers (in which case they will
get selected as potential reviewers), or emeritus_approvers (in
which case they've become inactive and would be poor choices for
reviewers)
out of 200 most recently updated merged PRs in 'kubernetes/kubernetes' involving 'andrewsykim'...
85 touched files matching regex '^test/'
of those, andrewsykim authored 8, and commented '/lgtm' on 34
I read through a number of reviews, they're substantive feedback
and not rubber stamps. Keep up the good work!
Originally, printOnce() was implemented for printing debugging message
once for checking actual value is the same as expected value.
However even if an error doesn't happen in ExpectEqual(), printOnce()
is called internally. Then the second ExpectEqual() doesn't output
the debugging message if an error happens. This makes debugging
difficult, let's just remove it for debugging.
