Refactor genericScheduler and signature of preemption funcs
- remove podNominator from genericScheduler
- simplify signature of preemption functions
Make Preempt() private
Previously, separate interfaces were defined for Reserve and Unreserve
plugins. However, in nearly all cases, a plugin that allocates a
resource using Reserve will likely want to register itself for Unreserve
as well in order to free the allocated resource at the end of a failed
scheduling/binding cycle. Having separate plugins for Reserve and
Unreserve also adds unnecessary config toil. To that end, this patch
aims to merge the two plugins into a single interface called a
ReservePlugin that requires implementing both the Reserve and Unreserve
methods.
and e2e_scheduling_duration_seconds
Also adding result label to e2e_scheduling_duration_seconds. Previously, the metric was only updated for successful attempts
Signed-off-by: Aldo Culquicondor <acondor@google.com>
- Add a defaultpreemption PostFilter plugin
- Make g.Preempt() stateless
- make g.Preempt() stateless
- make g.getLowerPriorityNominatedPods() stateless
- make g.processPreemptionWithExtenders() stateless
- replace error with NodeToStatusMap in Preempt() signature
- eliminate podPreemptor interface and expose its functions statelessly
- move logic in scheduler.go#preempt to generic_scheduler.go#Preempt()
- Make some private functions stateless
- make addNominatedPods() not dependent on genericScheduler
- make addNominatedPods() not dependent on genericScheduler
- make selectVictimsOnNode() not dependent on genericScheduler
- make selectNodesForPreemption() not dependent on genericScheduler
The scheduler doesn't really need to know in detail which reasons
rendered a node unusable for a node. All it needs from the volume
binder is a list of reasons that it then can present to the user.
This seems a bit cleaner. But the main reason for the change is that
it simplifies the checking of CSI inline volumes and perhaps later
capacity checking. Both will lead to new failure reasons, which then
can be added without changing the interface.
- eliminate running paths of Predicates in scheduler; use Filter Plugins instead.
- refactor all unit tests
- adjust the TestPreemptWithPermitPlugin integration test
Handle registry merge error.
Add test case
Add a test case in TestSchedulerCreation
Add test case comment
Add test case comment, and modify error msg.
gofmt
gofmt import order
replace NewDefaultRegistry with a fake registry
update comment
rename framework plugins nodeaffinity import
remove nodeaffinity plugin import, get fake registry name from default registry randomly
Add a space
Add a space after //
update comment, trans FrameWork to Framework
Changes scheduler.New so that algorithm source is moved from the
parameter to an option. The default algorithm source is source with the
DefaultProvider.
As part of graduating the scheduler's component config to beta, we require configurable fields to be nilable pointers (see https://github.com/kubernetes/kubernetes/issues/78109). This enables the ability to distinguish between default and unset values. We are only applying this change to external types, and reacting in our defaulting logic. This also reverts existing internal component config fields which were pointers to be non-pointers, for consistency.
- add Policy API to pkg/scheduler/apis/config and staging/src/k8s.io/kube-scheduler/config/v1
- dual-register Policy as apiGroup "v1" and "kubescheduler.config.k8s.io
- move/merge pkg/scheduler/api to pkg/scheduler/apis/config/...
- alias schedulerapi to pkg/scheduler/apis/config
- alias legacyapi to pkg/scheduler/api
- eliminate latest.Codec; use scheme.Codecs instead
- unit tests to verify Policy YAML with version "v1" or "kubescheduler.config.k8s.io/v1" can be loaded properly
- update api/api-rules/violation_exceptions.list
- move extender related API from pkg/scheduler/api to pkg/scheduler/apis/extender/v1
- alias extenderv1 to pkg/scheduler/apis/extender/v1
- use NodeScore and NodeScoreList in non-extender logic
The Configurator has been used as a holder for listers that tests need,
which is not its purpose. By making the tests obtain listers from more
appropriate places, such as informers, there is no need for various
accessors to the Configurator.
Also, FakeConfigurator is not being used anymore, so there's no need for
an interface instead of a plain pointer.
Signed-off-by: Aldo Culquicondor <acondor@google.com>
