This update introduces error handling for cases where a map of slices is encountered during validation. The validation process now explicitly returns an error message indicating that this type of validation is not supported, enhancing the robustness of the validation-gen tool.
Given the following:
```
type Struct type {
TypeMeta int
// +k8s:validateFalse="embedded"
OtherStruct
}
```
What should the field-path be for the error?
or:
```
type Struct type {
TypeMeta int
OtherStruct // embedded
}
// +k8s:validateFalse="otherstruct"
type OtherStruct {
I int
}
```
Conclusion: If we find an embedded field, we look for a nil fldPath, and
if so we set it to the type name. Embedded values in root types (e.g.
above `spec`) should not be a things we actually do.
Implementation: Add `safe.Value(*T, func() *T)`
kubeadm has replaced pkg/errors with an internal implementation, which
means that pkg/errors can be marked as unwanted (once a final
straggler is replaced).
Signed-off-by: Stephen Kitt <skitt@redhat.com>
This fixes the following issue in ci-benchmark-scheduler-perf-master:
malformed import path " ": invalid char ' '
FAIL [setup failed]
Setting the non-empty string with just a space was broken, the right way to
disable the race detector is with an empty KUBE_RACE variable. This worked
before when test-integration.sh unconditionally overwrote the KUBE_RACE
variable and stopped working when 6cb1488 started to keep whatever was
set by the caller.
This then caused an empty string to be passed as additional parameter to "go
test".
- Replace usage of pkg/errors with the local util/errors.
- Run kinder/hack/orderimports tool from kubeadm repo to
order the imports in the source tree.
pkg/errors is archived and while there is go-errors/errors as
an alternative, it lacks wraping methods. kubeadm has
specific neends and it's better to implement something
minimal locally instead of introducing another depedency.
- Implement basic wrapped errors and stack trace support.
cmd/kubeadm/app/util/errors.
- Remove unused error codes >1. At some point it seems we broke
these and 1 was returned for all error types.
- Remove the Error type in preflight and separate the printing
of '[preflight]' message and the error return from preflight
checks.
- Print an 'error:' prefix for all errors.
when a test is verifying a container has restarted, we use a continually exiting
container. Not verifying the number of restarts is less than (rather than equal) introduces
a race between the container restarting and the status observation.
Signed-off-by: Peter Hunt <pehunt@redhat.com>
With OrderedNamespaceDeletion enabled by default, the namespace
controller now deletes resources before deleting the namespace itself.
At the same time, some existing tests create many resources. When
those tests finish around the same time, the namespace controller can
become overloaded. As a result, it may take longer to start deleting
resources created by this test.
Initially, this patch increased the test timeout to account for that
delay. However, the test still flaked in 1 out of 10 jobs, which
suggests increasing the timeouts isn't enough.
Once the test was made serial, no flakes were observed.
in general, the rewritten e2e cpumanager test assume cgroup v2.
A limited set of these may be updated to work also with the
obsolete and declining cgroup v1, but these need to be reviewed
on test-by-test matter.
To fix test failures, we add a top level require for cgroup v2,
skipping otherwise. This will fix the red lanes while we review
the testcases and the deprecation plan of the other tests.
Signed-off-by: Francesco Romani <fromani@redhat.com>
