github/kubernetes
4
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
121,390 Commits 1 Branch 31 Tags 1,019 MiB
fcd6c19c24
Commit Graph

1473 Commits

Author SHA1 Message Date
AxeZhan
af26ebd0fa translate beta label to ga in volume_zone 2023-08-18 00:31:09 +08:00
SataQiu
ef7d404702 using wait.PollUntilContextTimeout instead of deprecated wait.Poll for pkg/scheduler 
using wait.PollUntilContextTimeout instead of deprecated wait.Poll for test/integration/scheduler

using wait.PollUntilContextTimeout instead of deprecated wait.Poll for test/e2e/scheduling

using wait.ConditionWithContextFunc for PodScheduled/PodIsGettingEvicted/PodScheduledIn/PodUnschedulable/PodSchedulingError
 2023-08-17 17:25:09 +08:00
kwakubiney
5752cbd8c7 chore: add pod UID in event log 
This change includes preemptor pod UID in event log to allow
for easier debugging.

Signed-off-by: kwakubiney <kebiney@hotmail.com>
 2023-08-16 11:00:56 +00:00
Kubernetes Prow Robot
130a5a423f
Merge pull request #119785 from sanposhiho/waitonpermit-fiterror 
fix: register the plugin rejects Pods in WaitOnPermit to UnschedulablePlugins
 2023-08-15 23:13:04 -07:00
Kubernetes Prow Robot
719d1a84f7
Merge pull request #119778 from sanposhiho/bugfix-unschedulableandunresolvable 
fix: when PreFilter returns UnschedulableAndUnresolvable, copy the state in all nodes in statusmap
 2023-08-15 23:12:57 -07:00
Kubernetes Prow Robot
57212647e9
Merge pull request #119769 from Huang-Wei/bug/prefilter-preemption 
Fix a bug that PostFilter plugin may don't function if previous PreFilter plugins return Skip
 2023-08-15 23:12:50 -07:00
Kubernetes Prow Robot
ea30d100f6
Merge pull request #119399 from wackxu/optimizecodeforNodeUnschedulable 
Optimize the code of NodeUnschedulable to reduce TolerationsTolerateT…
 2023-08-15 17:14:26 -07:00
AxeZhan
47fec59a31 parse node selector in prefilter 2023-08-14 16:39:46 +08:00
Kensei Nakada
cf3f0bd778 fix: register the plugin rejects Pods in WaitOnPermit to UnschedulablePlugins 2023-08-12 07:18:01 +00:00
Kensei Nakada
b008223705 fix: when PreFilter returns UnschedulableAndUnresolvable, copy the state in all nodes in statusmap 2023-08-12 06:58:49 +00:00
Wei Huang
765f3916c2
Fix a bug that PostFilter plugin may not function if previous PreFilter plugins return Skip 2023-08-10 13:43:00 -07:00
AxeZhan
2863b3d1ab Revert "refactor: simplify RunScorePlugins for readability + performance" 
This reverts commit a7eb7ed5c6.
 2023-07-20 10:50:32 +08:00
wackxu
a9d26ac7c7 Optimize the code of NodeUnschedulable to reduce TolerationsTolerateTaint function calls 
Signed-off-by: wackxu <xushiwei5@huawei.com>
 2023-07-18 21:00:05 +08:00
Adrian Cole
89ab733760 Improve docs on framework.CycleState 
Signed-off-by: Adrian Cole <adrian@tetrate.io>
Co-authored-by: Kante Yin <kerthcet@gmail.com>
 2023-07-18 14:48:20 +08:00
carlory
0599b3caa0 change the QueueingHintFn to pass a logger 2023-07-13 00:56:41 +08:00
Patrick Ohly
6f1a29520f scheduler/dra: reduce pod scheduling latency 
This is a combination of two related enhancements:
- By implementing a PreEnqueue check, the initial pod scheduling
  attempt for a pod with a claim template gets avoided when the claim
  does not exist yet.
- By implementing cluster event checks, only those pods get
  scheduled for which something changed, and they get scheduled
  immediately without delay.
 2023-07-12 11:17:04 +02:00
Patrick Ohly
ef48efc736 scheduler dynamicresources: minor logging improvements 
This makes some complex values a bit more readable.
 2023-07-12 11:07:59 +02:00
Kubernetes Prow Robot
e0dafe57a3
Merge pull request #117351 from pohly/dra-generated-resource-claim-names 
DRA: generated resource claim names
 2023-07-11 10:33:11 -07:00
Patrick Ohly
444d23bd2f dra: generated name for ResourceClaim from template 
Generating the name avoids all potential name collisions. It's not clear how
much of a problem that was because users can avoid them and the deterministic
names for generic ephemeral volumes have not led to reports from users. But
using generated names is not too hard either.

What makes it relatively easy is that the new pod.status.resourceClaimStatus
map stores the generated name for kubelet and node authorizer, i.e. the
information in the pod is sufficient to determine the name of the
ResourceClaim.

The resource claim controller becomes a bit more complex and now needs
permission to modify the pod status. The new failure scenario of "ResourceClaim
created, updating pod status fails" is handled with the help of a new special
"resource.kubernetes.io/pod-claim-name" annotation that together with the owner
reference identifies exactly for what a ResourceClaim was generated, so
updating the pod status can be retried for existing ResourceClaims.

The transition from deterministic names is handled with a special case for that
recovery code path: a ResourceClaim with no annotation and a name that follows
the Kubernetes <= 1.27 naming pattern is assumed to be generated for that pod
claim and gets added to the pod status.

There's no immediate need for it, but just in case that it may become relevant,
the name of the generated ResourceClaim may also be left unset to record that
no claim was needed. Components processing such a pod can skip whatever they
normally would do for the claim. To ensure that they do and also cover other
cases properly ("no known field is set", "must check ownership"),
resourceclaim.Name gets extended.
 2023-07-11 14:23:48 +02:00
Kubernetes Prow Robot
c95b16b280
Merge pull request #118608 from utam0k/podtopologyspread-prescore-skip 
Return Skip in PodTopologySpread#PreScore under specific conditions
 2023-07-10 09:27:07 -07:00
Kubernetes Prow Robot
0ae9aaacfa
Merge pull request #118271 from tangwz/add_nodeports_prefilter_skip_status 
feat(NodePorts): return Skip status in PreFilter
 2023-07-09 20:49:04 -07:00
Gunju Kim
7286d122fb
Mark pods with restartable init containers as UnschedulableAndUnresolvable 
This marks the pods with restartable init containers as
`UnschedulableAndUnresolvable` if the feature gate is disabled to avoid
the inconsistency in resource calculation between the scheduler and the
older kubelet.
 2023-07-08 07:26:13 +09:00
kerthcet
c0eb0caf4a Support fine-gained rescheduling in ReservePlugin 
Signed-off-by: kerthcet <kerthcet@gmail.com>
 2023-07-07 13:30:29 +08:00
Kubernetes Prow Robot
b07a843cb5
Merge pull request #119046 from kerthcet/fix/handle-unschedule-plugins 
Fix fitError in Permit plugin not handled perfectly
 2023-07-06 21:01:03 -07:00
kerthcet
278a8376e1 Fix: fiterror in permit plugin not handled perfectly 
We only added failed plulgins, but actually this will not work unless
we make the status with a fitError because we only copy the failured plugins
to podInfo if it is a fitError

Signed-off-by: kerthcet <kerthcet@gmail.com>
 2023-07-07 10:35:59 +08:00
Kubernetes Prow Robot
aeed7da616
Merge pull request #119077 from sanposhiho/follow-up-hint 
clean up the implementation around QueueingHintFn
 2023-07-06 13:39:15 -07:00
Kensei Nakada
be0db3f93d clean up the implementation around QueueingHintFn 2023-07-06 16:07:39 +00:00
tangwz
1bf2f6c9c0 feat(NodePorts): return Skip status in PreFilter 2023-07-06 08:42:08 +08:00
Kubernetes Prow Robot
293c1b8378
Merge pull request #118025 from AxeZhan/score-metrics 
feature(scheduler): plugin_evaluation_total metric support preScore/score
 2023-07-05 05:14:56 -07:00
Shingo Omura
d53762ec3a
remove unnecessary comment in pkg/scheduler/framework.QueueingHintFn 
event is not passed to QueueingHintFn but it exists a comment about it.
event is unnecessary in QueueingHintFn because QueueingHintFn is used in
ClusterEventWithHint and ClusterEventWithHint already have ClusterEvent.

Signed-off-by: Shingo Omura <everpeace@gmail.com>
 2023-06-29 21:22:20 +09:00
utam0k
ef26510164
Return Skip in PodTopologySpread#PreScore under specific conditions 
Signed-off-by: utam0k <k0ma@utam0k.jp>
 2023-06-28 12:08:10 +00:00
Kubernetes Prow Robot
52457842d1
Merge pull request #117055 from cyclinder/csi_migration 
remove CSI-migration gate
 2023-06-28 04:28:31 -07:00
Kubernetes Prow Robot
d9714078f8
Merge pull request #118551 from sanposhiho/event-to-register 
feature(scheduler): implement ClusterEventWithHint to filter out useless events
 2023-06-26 06:41:45 -07:00
Kensei Nakada
6f8d38406a feature(scheduler): implement ClusterEventWithHint to filter out useless events 2023-06-22 13:36:19 +00:00
Heba Elayoty
902c711fb4
Unset gated pod info timestamp in addToActiveQ 
Signed-off-by: Heba Elayoty <hebaelayoty@gmail.com>
 2023-06-21 14:16:08 -07:00
Kubernetes Prow Robot
bc8e312857
Merge pull request #117903 from sourcelliu/dynamic 
feature(DynamicResources): return Skip in PreFilter
 2023-06-20 17:48:20 -07:00
Kubernetes Prow Robot
9740bc0e0a
Merge pull request #118606 from sanposhiho/refactor-score 
refactor: simplify RunScorePlugins for readability + performance
 2023-06-13 21:41:57 -07:00
Kensei Nakada
a7eb7ed5c6 refactor: simplify RunScorePlugins for readability + performance 2023-06-11 03:29:05 +00:00
Kubernetes Prow Robot
4483bf66fe
Merge pull request #116635 from mengjiao-liu/contextual-logging-plugin-interpodaffinity 
Migrated `pkg/scheduler/framework/plugins/interpodaffinity` to contextual logging
 2023-06-09 08:14:13 -07:00
guoguangwu
1d9eed9f95 chore: slice replace loop 2023-06-05 22:40:53 +08:00
SataQiu
410b6023d6 scheduler: fix code style issues for pkg/scheduler 2023-06-05 17:29:49 +08:00
cyclinder
8e4228a8c1 remove CSI-migration gate 2023-06-04 18:40:17 +08:00
Mengjiao Liu
6d23da045f Migrated pkg/scheduler/framework/plugins/interpodaffinity to use contextual logging 2023-06-01 18:24:54 +08:00
Mengjiao Liu
074900e81b scheduler: update the scheduler interface and cache methods to use contextual logging 2023-05-29 13:26:32 +08:00
Kubernetes Prow Robot
f7cfb5f02f
Merge pull request #118257 from pohly/dra-scheduler-plugin-loopvar-fix 
dra scheduler plugin test: fix loopvar bug and "reserve" expected data
 2023-05-26 06:06:53 -07:00
Patrick Ohly
7a6b4a9215 dra scheduler plugin test: fix loopvar bug and "reserve" expected data 
The `listAll` function returned a slice where all pointers referred to the same
instance. That instance had the value of the last list entry. As a result, unit
tests only compared that element.

During the reserve phase, the first claim gets reserved in two test
cases. Those two tests must expect that change. That hadn't been noticed before
because that first claim didn't get compared.
 2023-05-25 15:10:05 +02:00
Mengjiao Liu
1c05cf1d51 kube-scheduler: NewFramework function to pass the context parameter 
Co-authored-by: Aldo Culquicondor <1299064+alculquicondor@users.noreply.github.com>
 2023-05-23 10:17:34 +08:00
Kubernetes Prow Robot
c7c41d27b4
Merge pull request #117834 from NoicFank/cleanup-scheduler-node-must-not-nil-in-snapshot 
cleanup useless null pointer check about nodeInfo.Node() from snapshot for in-tree plugins
 2023-05-20 15:16:18 -07:00
dingzhu lurong
ed26fcf5b8 cleanup useless null pointer check about nodeInfo.Node() from snapshot for in-tree plugins 2023-05-20 22:53:43 +08:00
Kubernetes Prow Robot
da1b9df26c
Merge pull request #118032 from kerthcet/cleanup/interpodaffinity2 
Chore: cleanup in interPodAffinity
 2023-05-17 14:00:33 -07:00
Kubernetes Prow Robot
53772982be
Merge pull request #116829 from mengjiao-liu/contextual-logging-scheduler-plugin-volumezone 
Migrated the volumezone scheduler plugin to use contextual logging
 2023-05-16 09:53:35 -07:00
AxeZhan
8c6c1e5319 plugin_evaluation_total metric support preScore/score 2023-05-16 16:52:29 +08:00
kerthcet
3ac7497361 Chore: cleanup in interpodaffinity 
Signed-off-by: kerthcet <kerthcet@gmail.com>
 2023-05-16 14:46:15 +08:00
mantuliu
6e2ea32fc8 feature(DynamicResources): return Skip in PreFilter 2023-05-15 00:06:08 +08:00
Kubernetes Prow Robot
58e13496d6
Merge pull request #116842 from mengjiao-liu/contextual-logging-scheduler-runtime 
Migrated `pkg/scheduler/framework/runtime` to use contextual logging
 2023-05-11 10:59:02 -07:00
Mengjiao Liu
fe728996ca scheduler test: call frameworkruntime.WithLogger function for contextual logging 2023-05-11 15:46:08 +08:00
Mengjiao Liu
b456392567 Migrated pkg/scheduler/framework/runtime to use contextual logging 2023-05-11 15:46:02 +08:00
utam0k
c0611b6bb3
Return Skip in InterPodAffinity#PreScore under specific conditions 
This commit updates the InterPodAffinity PreScore to return a Skip status when the following conditions are met:
1. There are no nodes to score.
2. The incoming pod has no inter-pod affinities && the `IgnorePreferredTermsOfExistingPods` option is enabled.

Signed-off-by: utam0k <k0ma@utam0k.jp>
 2023-05-10 13:02:23 +00:00
mantuliu
e6900f5ead Optimize the performance of the Clone method of preFilterState 2023-05-05 19:53:12 +08:00
mantuliu
887654160f Improve the performance of schedulinggates 2023-05-04 18:09:49 +08:00
Kubernetes Prow Robot
47f1bd9f80
Merge pull request #117649 from SataQiu/scheduler-remove-v1beta2-20230427 
scheduler: remove deprecated v1beta2 KubeSchedulerConfiguration  component config
 2023-05-03 09:54:41 -07:00
Kubernetes Prow Robot
0d67dd689b
Merge pull request #117683 from utam0k/skip-topologyspread-empty 
Add check to skip PodTopologySpread PreFilter if no constraints are specified
 2023-05-03 06:48:24 -07:00
SataQiu
1f7c07f355 scheduler: remove deprecated v1beta2 KubeSchedulerConfiguration 2023-05-03 21:43:19 +08:00
utam0k
d82684e691
Add check to skip PodTopologySpread PreFilter if no constraints are specified 
This commit adds a check in the PodTopologySpread PreFilter function to
return a Skip status if there are no topology spread constraints specified
This prevents unnecessary processing and filtering for pods that don't have any topology spread
constraints.
This change is a part of the work for issue #114399.

Signed-off-by: utam0k <k0ma@utam0k.jp>
 2023-05-03 04:39:00 +00:00
Kubernetes Prow Robot
8353d4623b
Merge pull request #117427 from cbandy/pkg-testing-setenv 
Replace os.Setenv with testing.T.Setenv in tests
 2023-04-29 08:28:16 -07:00
Kubernetes Prow Robot
b44482a37c
Merge pull request #116797 from mengjiao-liu/contextual-looging-scheduler-plugin-podtopologyspread 
Migrated `pkg/scheduler/framework/plugins/podtopologyspread` to contextual logging
 2023-04-27 12:28:27 -07:00
Kubernetes Prow Robot
a38efaccc0
Merge pull request #116748 from mengjiao-liu/contextual-logging-scheduler-plugin-noderesource 
Migrated `pkg/scheduler/framework/plugins/noderesources` to contextual logging
 2023-04-27 12:28:15 -07:00
Kubernetes Prow Robot
5170c25609
Merge pull request #116835 from mengjiao-liu/contextual-logging-scheduler-plugin-preemption 
Migrated `pkg/scheduler/framework/preemption & defaultpreemption` to use contextual logging
 2023-04-27 11:10:16 -07:00
Kubernetes Prow Robot
87f3acf7f6
Merge pull request #115398 from tangwz/add_NodeVolumeLimits_PreFilter 
feat(NodeVolumeLimits): return Skip in PreFilter
 2023-04-27 01:44:14 -07:00
Mengjiao Liu
7f370d651d Migrated pkg/scheduler/framework/plugins/podtopologyspread to contextual logging 2023-04-27 15:55:09 +08:00
Mengjiao Liu
54e6f609ce Migrated pkg/scheduler/framework/plugins/noderesources to contextual logging 2023-04-27 14:46:13 +08:00
Mengjiao Liu
37a9260d5c Migrate pkg/scheduler/framework/plugins/defaultpreemption/default_preemption.go to use contextual logging 2023-04-27 11:28:19 +08:00
Mengjiao Liu
eeb1399383 Migrated pkg/scheduler/framework/preemption to use contextual logging 2023-04-27 11:28:14 +08:00
tangwz
8ed861889a feat(NodeVolumeLimits): return Skip in PreFilter 2023-04-26 20:17:04 +08:00
Antoni Zawodny
28519e240e Fix scheduler performance regression after adding plugin metrics 2023-04-26 14:11:48 +02:00
Maciej Borsz
5c584269a7 avoid volume copy in checkAttachableInlineVolume 2023-04-19 20:10:22 +00:00
Chris Bandy
d38ac7e7c6 Replace os.Setenv with testing.T.Setenv in tests 
T.Setenv ensures that the environment is returned to its prior state
when the test ends. It also panics when called from a parallel test to
prevent racy test interdependencies.
 2023-04-17 20:39:46 -05:00
Kubernetes Prow Robot
94a15929cf
Merge pull request #116408 from ChenLingPeng/fit 
skip pod resource check when request is zero
 2023-04-17 11:44:45 -07:00
Kubernetes Prow Robot
af99aa1a50
Merge pull request #117214 from HirazawaUi/add-preemptor-log 
add log includes pod preemption details
 2023-04-13 12:40:38 -07:00
HirazawaUi
ed0124855a add log includes pod preemption details 2023-04-13 22:29:18 +08:00
Kubernetes Prow Robot
242702cb86
Merge pull request #116940 from sarab97/sarab/feat/sets 
Use the generic Set in scheduler
 2023-04-11 19:17:14 -07:00
Kubernetes Prow Robot
365ac69fc4
Merge pull request #116845 from major1201/fix_binder_typo 
fix GetPodVolumeClaims in comments
 2023-04-11 18:20:05 -07:00
Kubernetes Prow Robot
e77ca49022
Merge pull request #114898 from AxeZhan/volumerestrictions 
feature(volume_restrictions): return Skip in PreFilter
 2023-04-11 15:35:04 -07:00
sarab
8d18ae6fc2 Use the generic Set in scheduler 2023-04-09 11:34:17 +05:30
kidddddddddddddddddddddd
8d644fbc72 return skip in volumerestrictions 2023-03-23 23:14:24 +08:00
major1201
daa1d287e4 fix GetPodVolumeClaims in comments 2023-03-22 17:23:05 +08:00
Mengjiao Liu
27f84b755e Migrated the volumezone scheduler plugin to use contextual logging 2023-03-22 11:09:31 +08:00
Kubernetes Prow Robot
0e8eae6e35
Merge pull request #116607 from alculquicondor/optimize-spreading 
Optimize topology spreading for null selector
 2023-03-14 15:14:57 -07:00
Kubernetes Prow Robot
fbfc887a09
Merge pull request #116556 from pohly/dra-podschedulingcontext 
dra: PodScheduling -> PodSchedulingContext
 2023-03-14 15:14:34 -07:00
Kubernetes Prow Robot
f2e1a67c05
Merge pull request #116535 from denkensk/fix-match 
feat: forbid to set matchLabelKeys when labelSelector isn’t set in topologySpreadConstraints
 2023-03-14 14:13:04 -07:00
Aldo Culquicondor
3dae9ba6d6
Optimize topology spreading for null selector 
Change-Id: I28f031a040b143fa9452b4400cc3ae85d58e572c
 2023-03-14 15:01:59 -04:00
Kubernetes Prow Robot
25edb4943a
Merge pull request #116465 from sourcelliu/nodeaffinity 
Fix PreFilter logic problem in node affinity
 2023-03-14 10:40:52 -07:00
Kubernetes Prow Robot
49649c89ea
Merge pull request #113584 from yangjunmyfm192085/volume-contextual-logging 
volume: use contextual logging
 2023-03-14 10:40:16 -07:00
Kubernetes Prow Robot
e660a1702b
Merge pull request #116571 from mengjiao-liu/scheduler-plugins-defaultbinder 
Migrate pkg/scheduler/framework/plugins/defaultbinder to contextual logging
 2023-03-14 09:39:59 -07:00
Patrick Ohly
fec5233668 api: resource.k8s.io PodScheduling -> PodSchedulingContext 
The name "PodScheduling" was unusual because in contrast to most other names,
it was impossible to put an article in front of it. Now PodSchedulingContext is
used instead.
 2023-03-14 10:18:08 +01:00
Mengjiao Liu
c512232782 Migrate pkg/scheduler/framework/plugins/defaultbinder to contextual logging 2023-03-14 17:02:25 +08:00
Patrick Ohly
29941b8d3e api: resource.k8s.io v1alpha1 -> v1alpha2 
For Kubernetes 1.27, we intend to make some breaking API changes:
- rename PodScheduling -> PodSchedulingHints (https://github.com/kubernetes/kubernetes/issues/114283)
- extend ResourceClaimStatus (https://github.com/kubernetes/enhancements/pull/3802)

We need to switch from v1alpha1 to v1alpha2 for that.
 2023-03-14 07:52:03 +01:00
Kubernetes Prow Robot
0701e4b3b0
Merge pull request #116508 from tangwz/fix_PreFilter_and_PreScore_test_name 
Avoid using negative words in PreFilter and PreScore tests.
 2023-03-13 21:17:20 -07:00
Kubernetes Prow Robot
e3cf864279
Merge pull request #116502 from sourcelliu/removecast 
Remove unnecessary int64 type conversion
 2023-03-13 21:17:08 -07:00
Alex Wang
08dfa41f17 fix: key in matchLabelKeys needs to be ignored when LabelSelector is nil 
Signed-off-by: Alex Wang <wangqingcan1990@gmail.com>
 2023-03-14 11:25:04 +08:00
mantuliu
42ddbb4276 Fix PreFilter logic problem in node affinity 2023-03-14 11:23:40 +08:00
杨军10092085
361e4ff0fa volume: use contextual logging 2023-03-14 08:37:30 +08:00
Kubernetes Prow Robot
3106a5c553
Merge pull request #116301 from andyzhangx/remove-azuredisk-code 
Remove Azure disk in-tree storage plugin
 2023-03-13 10:38:48 -07:00
Kubernetes Prow Robot
3c6e419cc3
Merge pull request #116450 from vinaykul/restart-free-pod-vertical-scaling-api 
Rename ContainerStatus.ResourcesAllocated to ContainerStatus.AllocatedResources
 2023-03-12 16:06:40 -07:00
Kubernetes Prow Robot
a32050e6cb
Merge pull request #116201 from sanposhiho/metric-scheduling-gate 
add(scheduler): implement "plugin_execution_duration_seconds" metric in PreEnqueue
 2023-03-12 13:52:40 -07:00
Kubernetes Prow Robot
e413e6a59c
Merge pull request #116312 from sanposhiho/move-label 
cleanup(scheduler): move metric labels to metrics package
 2023-03-12 10:48:40 -07:00
tangwz
3766e060e5 Avoid using negative words in PreFilter and PreScore tests. 2023-03-12 15:06:26 +08:00
Kensei Nakada
639007b28e cleanup(scheduler): move metric labels to metrics package 2023-03-12 05:10:29 +00:00
Kensei Nakada
6697467062 add(scheduler): implement "plugin_execution_duration_seconds" metric in PreEnqueue 2023-03-12 04:45:52 +00:00
mantuliu
4204251acf Remove unnecessary int64 type conversion 
Signed-off-by: mantuliu <240951888@qq.com>
 2023-03-12 09:50:09 +08:00
tangwz
be080584c6
scheduler(NodeResourcesFit & NodeResourcesBalancedAllocation): calculatePodResourceRequest in PreScore phase (#115655) 
* scheduler(NodeResourcesFit): calculatePodResourceRequest in PreScore phase

* scheduler(NodeResourcesFit and NodeResourcesBalancedAllocation): calculatePodResourceRequest in PreScore phase

* modify the comments and tests.

* revert the tests.

* don't need consider nodes.

* use list instead of map.

* add comment for podRequests.

* avoid using negative wording in variable names.
 2023-03-10 07:44:53 -08:00
vinay kulkarni
01b96e7704 Rename ContainerStatus.ResourcesAllocated to ContainerStatus.AllocatedResources 2023-03-10 14:49:26 +00:00
Kubernetes Prow Robot
8b3d529523
Merge pull request #116395 from alculquicondor/fix-podinfo-race 
One lock among PodNominator and SchedulingQueue
 2023-03-09 22:44:17 -08:00
Todd Neal
4096c9209c dedupe pod resource request calculation 2023-03-09 17:15:53 -06:00
Maciej Borsz
30bca1e1d5 Avoid metric lookup in Parallelizer.Util on every work piece 2023-03-09 17:12:30 +00:00
andyzhangx
5d0a54dcb5 remove Azure Disk in-tree driver code 
fix
 2023-03-09 13:24:08 +00:00
forrestchen
bbf2b968c8 skip pod resource check when request is zero 
Signed-off-by: forrestchen <forrestchen@tencent.com>
 2023-03-09 17:05:11 +08:00
Aldo Culquicondor
07a73bb2e1
One lock among PodNominator and SchedulingQueue 
Change-Id: I17fe5da40250e42c04124c25b530ce6c8dea4154
 2023-03-08 16:18:36 -05:00
Kubernetes Prow Robot
83334ccaa1
Merge pull request #116320 from wangchen615/lastminute-scheduler-fix 
Address last-minute requested changes for inplace update feature testing in scheduler
 2023-03-07 19:31:26 -08:00
Kubernetes Prow Robot
9d7db70882
Merge pull request #116066 from yoongon/feature/fit-test-comments 
Fix comments at fit_test.go to increase redability
 2023-03-06 20:52:16 -08:00
Chen Wang
fd6105d015 fix last minute scheduler changes for inplace update 2023-03-06 18:47:02 -05:00
Kubernetes Prow Robot
283c26f91a
Merge pull request #115082 from sanposhiho/filter-metrics 
feature(scheduler): implement plugin_evaluation_total metric
 2023-03-06 12:58:17 -08:00
Kubernetes Prow Robot
89d1a7971e
Merge pull request #116287 from csDengh/minor_code_improvement 
scheduler: get an element from the nodes slice out of the loop for the better performance
 2023-03-06 09:24:49 -08:00
Kubernetes Prow Robot
d6e9cff212
Merge pull request #115838 from torredil/remove-aws 
Remove AWS legacy cloud provider + EBS in-tree storage plugin
 2023-03-06 08:18:29 -08:00
torredil
6aebda9b1e Remove AWS legacy cloud provider + EBS in-tree storage plugin 
Signed-off-by: torredil <torredil@amazon.com>
 2023-03-06 14:01:15 +00:00
csDengh
f762145e06
minor code improvement 
minor code improvement 
from repeated assignments in loops to initialize outside the loop
 2023-03-06 09:00:40 +08:00
Kensei Nakada
608f4808ff support PreFilter as well 2023-03-06 00:48:30 +00:00
mantuliu
83fdbd76a1 Improve the performance when Resource Clone 
Signed-off-by: mantuliu <240951888@qq.com>
 2023-03-05 00:35:51 +08:00
Yoon Park
8d2c81e7ec Fix comments at fit_test.go to increase redability 2023-03-04 13:03:15 +09:00
Kubernetes Prow Robot
b9fd1802ba
Merge pull request #102884 from vinaykul/restart-free-pod-vertical-scaling 
In-place Pod Vertical Scaling feature
 2023-02-27 22:53:15 -08:00
Kubernetes Prow Robot
e6522f8053
Merge pull request #116095 from sourcelliu/preemption 
Improve the performance of method CandidatesToVictimsMap
 2023-02-27 19:25:15 -08:00
mantuliu
ccde866a2f Improve the performance of method CandidatesToVictimsMap 
Signed-off-by: mantuliu <240951888@qq.com>
 2023-02-27 23:53:19 +08:00
moriya
6b4d08bd5c fix: typo 2023-02-25 19:06:07 +09:00
vinay kulkarni
231849a908 In-place Pod Vertical Scaling - Scheduler changes 2023-02-24 18:14:21 +00:00
TommyStarK
5531295bf9 scheduler/framework/plugins/volumebinding: fix inaccurate log for when a volume is bound to a claim 
Signed-off-by: TommyStarK <thomasmilox@gmail.com>
 2023-02-24 16:43:38 +01:00
Kubernetes Prow Robot
4dfd2fe74d
Merge pull request #115922 from sourcelliu/impovevolume 
Improve performance of method haveOverlap
 2023-02-21 11:34:08 -08:00
mantuliu
edc0455c5c Improve performance of method haveOverlap 
Signed-off-by: mantuliu <240951888@qq.com>
 2023-02-22 00:51:55 +08:00
mantuliu
af85f410a2 Improve the performance of slice usage 
Signed-off-by: mantuliu <240951888@qq.com>
 2023-02-21 19:30:27 +08:00
mantuliu
e1e07e34e3 Improve performance of method haveOverlap 
Signed-off-by: mantuliu <240951888@qq.com>
 2023-02-21 16:50:10 +08:00
Yoon Park
c6276cc2ab Fix typo at most_allocated_test.go 2023-02-21 10:32:40 +09:00
Kubernetes Prow Robot
77af0be42f
Merge pull request #115519 from sanposhiho/move-metric-recorder 
cleanup(scheduler): move metricRecorder to metrics package
 2023-02-16 03:57:38 -08:00
Kensei Nakada
aa7b1766e5 cleanup(scheduler): move metricRecorder to metrics package 2023-02-16 07:27:08 +00:00
Kubernetes Prow Robot
fdc117a6dd
Merge pull request #114887 from binacs/binacs/simplify-pickOneNodeForPreemption 
cleanup: simplify pickOneNodeForPreemption in preemption
 2023-02-14 14:25:51 -08:00
Kubernetes Prow Robot
a0878eb715
Merge pull request #115709 from yoongon/feature/default-binder-testnode 
Use predefined variable instead of duplicated string
 2023-02-13 14:53:49 -08:00
kidddddddddddddddddddddd
f5a69ffda9
feature(scheduler): won't run Score if PreScore returned a Skip status (#115652) 
* allow preScore to return skip status to skip running the corresponding score extension

* add test case for all skipped

* add test case for select host

* update plugin status

* skip score when all plugins are skipped

* update
 2023-02-13 14:53:29 -08:00
Yoon
a3405f35b4 Use predefined variable instead of duplicated string 2023-02-12 20:50:39 +09:00
Kensei Nakada
4f7717842c feature(scheduler): implement plugin_evaluation_total metric 2023-02-04 07:21:03 +00:00
Chris Henzie
dbc7d8ded0 feat: support preemption for pods using ReadWriteOncePod PVCs 
PVCs using the ReadWriteOncePod access mode can only be referenced by a
single pod. When a pod is scheduled that uses a ReadWriteOncePod PVC,
return "Unschedulable" if the PVC is already in-use in the cluster.

To support preemption, the "VolumeRestrictions" scheduler plugin
computes cycle state during the PreFilter phase. This cycle state
contains the number of references to the ReadWriteOncePod PVCs used by
the pod-to-be-scheduled.

During scheduler simulation (AddPod and RemovePod), we add and remove
reference counts from the cycle state if they use any of these
ReadWriteOncePod PVCs.

In the Filter phase, the scheduler checks if there are any PVC reference
conflicts, and returns "Unschedulable" if there is a conflict.

This is a required feature for the ReadWriteOncePod beta. See for more context:
https://github.com/kubernetes/enhancements/tree/master/keps/sig-storage/2485-read-write-once-pod-pv-access-mode#beta
 2023-01-30 10:59:22 -08:00
BinacsLee
b3af95cb8a cleanup: simplify pickOneNodeForPreemption in preemption 2023-01-29 19:55:43 +08:00
Sunny Song
5e2f12e943 Fix nil pointer error in nodevolumelimits csi logging 2023-01-25 09:37:46 -08:00
Patrick Ohly
bc6c7fa912 logging: fix names of keys 
The stricter checking with the upcoming logcheck v0.4.1 pointed out these names
which don't comply with our recommendations in
https://github.com/kubernetes/community/blob/master/contributors/devel/sig-instrumentation/migration-to-structured-logging.md#name-arguments.
 2023-01-23 14:24:29 +01:00
Kubernetes Prow Robot
7b1b8d5734
Merge pull request #114889 from sanposhiho/skip/inter-pod-affinity 
feature(InterPodAffinity): return Skip in PreFilter
 2023-01-18 20:36:35 -08:00
Kensei Nakada
3b20a007ed feature(InterPodAffinity): return Skip in PreFilter 2023-01-19 01:24:18 +00:00
Kubernetes Prow Robot
98da10e4f2
Merge pull request #115052 from kidddddddddddddddddddddd/fix/volume_zone 
fix volume_zone, iterate all topologyLabels when check node
 2023-01-18 10:54:46 -08:00
Kubernetes Prow Robot
f16e62ae0c
Merge pull request #114897 from kidddddddddddddddddddddd/volumebinding 
feature(volumebinding): return Skip in PreFilter
 2023-01-16 23:38:32 -08:00
kidddddddddddddddddddddd
2ba4c00b3c hasAnyNodeConstraint 2023-01-14 15:29:54 +08:00
Kubernetes Prow Robot
eabb70833a
Merge pull request #114896 from kidddddddddddddddddddddd/volume_zone 
feature(volume_zone): return Skip in PreFilter
 2023-01-13 23:14:13 -08:00
Kubernetes Prow Robot
3a8e2e399b
Merge pull request #114393 from danielvegamyhre/myfeature 
Option to ignore existing pods' preferred inter-pod affinities if the incoming pod has no preferred inter-pod affinities
 2023-01-13 17:28:15 -08:00
Daniel Vega-Myhre
41817b1888 optionally ignore preferred terms of existing pods unless incoming pod 
has inter-pod affinities
 2023-01-13 23:15:53 +00:00
Michal Wozniak
437179afc3 Do not include scheduler name in the preemption event message 2023-01-11 09:32:21 +01:00
Kubernetes Prow Robot
1d6ae20301
Merge pull request #114798 from kerthcet/cleanup/code-refactor 
Code refactor for readability in `RunFilterPlugins`
 2023-01-09 17:31:12 -08:00
Kubernetes Prow Robot
1e3946ce9d
Merge pull request #114923 from mimowo/do-not-leak-pod-name-in-event 
Adjust preemption event message to do not include preemptor pod metadata
 2023-01-09 13:51:28 -08:00
Kubernetes Prow Robot
eb7fd7f51c
Merge pull request #114914 from mimowo/do-not-leak-pod-name 
Adjust DisruptionTarget condition message to do not include preemptor pod metadata
 2023-01-09 11:15:40 -08:00
Michal Wozniak
f79a34d267 Do not leak cross namespace pod metadata in preemption events 2023-01-09 18:30:19 +01:00
Michal Wozniak
bdf58ce2eb Adjust DisruptionTarget condition message to do not include preemptor metadata 2023-01-09 12:22:19 +01:00
kidddddddddddddddddddddd
733d5695f2 always run filter in test 2023-01-08 11:13:16 +08:00
kidddddddddddddddddddddd
5411c05460 save state data for reserve 2023-01-08 10:49:35 +08:00
kidddddddddddddddddddddd
0abdf6abc2 revert check in filter 2023-01-07 22:30:16 +08:00
kidddddddddddddddddddddd
059d520537 return skip 2023-01-07 21:58:54 +08:00
kidddddddddddddddddddddd
de7c8db7cb return skip 2023-01-07 21:48:30 +08:00
Kensei Nakada
570c2d7036 cleanup(nodeaffinity): remove impossible scenario from test cases 2023-01-07 08:46:35 +00:00
Kante Yin
2ceadfe885 Code refactor for readability 
Signed-off-by: Kante Yin <kerthcet@gmail.com>
 2023-01-07 11:31:46 +08:00
Kubernetes Prow Robot
c549b59983
Merge pull request #114125 from sanposhiho/skip-reimplementation 
feature(scheduler): won't run Filter if PreFilter returned a Skip status
 2023-01-06 02:25:59 -08:00
Kubernetes Prow Robot
bd43394467
Merge pull request #114623 from Huang-Wei/feat/smart-preemption-identification 
Enhanced logic to identify eligible preemption node
 2023-01-06 00:23:59 -08:00
Kensei Nakada
e5b536755f feature(scheduler): won't run Filter if PreFilter returned a Skip status 2023-01-06 07:38:15 +00:00
Wei Huang
9b64025f36
rename 'PreemptionByKubeScheduler' to 'PreemptionByScheduler' 2023-01-05 10:31:57 -08:00
Wei Huang
91742e2393
Enhanced logic to identify eligible preemption node 2023-01-05 10:31:57 -08:00
Kante Yin
2c205e291d Distinguish unschedulable with unresolvable in scheduler 
Before, in RunPostFilterPlugins, we didn't distinguish between unschedulable and unresolvable
because we only have one postFilterPlugin by default, now, we have at least two, we should
make sure that once a postFilterPlugin returns unresolvable, we'll return directly

Signed-off-by: Kante Yin <kerthcet@gmail.com>
 2023-01-05 16:59:19 +08:00
Kante Yin
49e7e80999 Modify the return type of RunFilterPlugins to *Status 
Before, the return type of RunFilterPlugins is a Map, but considering we'll return immediately
once we met unsuccessful status, this is not necessary.

Signed-off-by: Kante Yin <kerthcet@gmail.com>
 2023-01-03 14:33:58 +08:00
Kubernetes Prow Robot
3a5829044c
Merge pull request #114098 from kidddddddddddddddddddddd/feat/pv_list 
Optimize volumebinding by moving PV list calls to PreFilter
 2022-12-22 10:11:37 -08:00
Kubernetes Prow Robot
419e0ec3d2
Merge pull request #114390 from tangwz/improve_NodeResourcesFit_replace_small_maps_with_slices 
Improve performance of NodeResourcesFit scoring
 2022-12-22 06:57:26 -08:00
kidddddddddddddddddddddd
27c5c0a16d pv_list to prefilter 2022-12-22 14:59:41 +08:00
Kubernetes Prow Robot
70c28f3e12
Merge pull request #114486 from kerthcet/cleanup/make-preemption-more-readable 
Make handling scheduleResult more readable
 2022-12-21 15:01:25 -08:00
Kante Yin
c8908716ee Make handling scheduleResult more readable 
Signed-off-by: Kante Yin <kerthcet@gmail.com>
 2022-12-22 01:22:34 +08:00
tangwz
a4ab559ba4 Improve performance of NodeResourcesFit scoring 
Signed-off-by: tangwz <tangwz.com@gmail.com>
 2022-12-21 23:23:47 +08:00
Kubernetes Prow Robot
a3ca5fd2c8
Merge pull request #114456 from kidddddddddddddddddddddd/status_with_error 
return error message when calling status.Message()
 2022-12-19 11:09:43 -08:00
kidddddddddddddddddddddd
e789beb213 errMsg 2022-12-19 23:52:06 +08:00
Jordan Liggitt
78cb3862f1
Fix indentation/spacing in comments to render correctly in godoc 2022-12-17 23:27:38 -05:00
Kubernetes Prow Robot
94fd106a17
Merge pull request #110453 from claudiubelu/unittests-4 
unittests: Fixes unit tests for Windows (part 4)
 2022-12-16 06:16:16 -08:00
Wei Huang
0aecf6b3d5
Fix a regression that scheduler always go through all Filter plugins 
The correct behavior is to stop upon the first Filter failure.
 2022-12-15 12:02:27 -08:00
Kubernetes Prow Robot
c2b5457dfa
Merge pull request #109877 from yibozhuang/local-pv-prefilter-result 
scheduler volumebinding: leverage PreFilterResult for bound local PVs
 2022-12-13 14:53:34 -08:00
Kubernetes Prow Robot
dc1e77143f
Merge pull request #114082 from kidddddddddddddddddddddd/refactor_handleSchedulingFailure 
pass status to handleSchedulingFailure
 2022-12-12 22:05:34 -08:00
kidddddddddddddddddddddd
6ca62eb2cb refactor 2022-12-13 11:36:12 +08:00
Kubernetes Prow Robot
c80b03fee8
Merge pull request #113499 from kerthcet/cleanup/map-empty-judgement 
Fix: resourceToWeightMap will never be nil here
 2022-12-12 00:15:33 -08:00
Kubernetes Prow Robot
a800f1e0c0
Merge pull request #114216 from alculquicondor/avoid-serialization 
Avoid serialization of maps when log level < 10
 2022-12-10 07:54:24 -08:00
kidddddddddddddddddddddd
9e0a7c7ee7 add pre_filter for volume_zone 2022-12-03 02:11:04 +08:00
Aldo Culquicondor
b62541fe84
Avoid serialization of maps when log level < 10 
Change-Id: If264c26ae599602d5b4919301f011c2095d706ee
 2022-11-30 12:53:52 -05:00
Claudiu Belu
4dc7a260ec unittests: Fixes unit tests for Windows (part 4) 
Currently, there are some unit tests that are failing on Windows due to
various reasons:

- paths not properly joined (filepath.Join should be used).
- files not closed, which means that they cannot be removed / renamed.
- time.Now() is not as precise on Windows, which means that 2
  consecutive calls may return the same timestamp.
 2022-11-29 16:32:26 +00:00
Wei Huang
a731a44596
Fix an accuracy issue of scheduler_pending_pods metric 2022-11-21 21:33:16 -08:00
Yibo Zhuang
380c7f248e scheduler volumebinding: leverage PreFilterResult 
This change will leverage the new PreFilterResult
to reduce down the list of eligible nodes for pod
using Bound Local PVs during PreFilter stage so
that only the node(s) which local PV node affinity
matches will be cosnidered in subsequent scheduling
stages.

Today, the NodeAffinity check is done during Filter
which means all nodes will be considered even though
there may be a large number of nodes that are not
eligible due to not matching the pod's bound local
PV(s)' node affinity requirement. Here we can
reduce down the node list in PreFilter to ensure that
during Filter we are only considering the reduced
list and thus can provide a more clear message to
users when node(s) are not available for scheduling
since the list only contains relevant nodes.

If error is encountered (e.g. PV cache read error) or
if node list reduction cannot be done (e.g. pod uses
no local PVs), then we will still proceed to consider
all nodes for the rest of scheduling stages.

Signed-off-by: Yibo Zhuang <yibzhuang@gmail.com>
 2022-11-17 21:37:24 -08:00