github/kubernetes
1
0
Fork 0
mirror of https://github.com/k3s-io/kubernetes.git synced 2026-07-23 10:26:22 +00:00
Code Issues Projects Releases Wiki Activity
126,221 Commits 42 Branches 9,895 Tags
9d1b0654e0b9cd4cd3d99f33019a8b00fc65cbdf
Commit Graph

51078 Commits

Author SHA1 Message Date
Patrick Ohly
9d1b0654e0 DRA: add wg/device-management label automatically 
This makes PRs show up automatically in the WG's project
board (https://github.com/orgs/kubernetes/projects/95/views/1).
 2024-10-28 16:36:04 +01:00
Kubernetes Prow Robot
3690cb7f9a Merge pull request #128101 from pohly/dra-api-cel-cost-limit 
DRA API: implement CEL cost limit
 2024-10-26 20:18:52 +01:00
Kubernetes Prow Robot
25d6f76538 Merge pull request #128337 from torredil/fix-gce-cos-master-serial-5123 
Add VolumeAttachment event registration to CSI volume limits plugin
 2024-10-26 16:00:52 +01:00
torredil
fe1badf635 Add VolumeAttachment event registration to CSI volume limits plugin 
Signed-off-by: torredil <torredil@amazon.com>
 2024-10-26 13:41:28 +00:00
Kubernetes Prow Robot
b3cf9c6e5c Merge pull request #128269 from tallclair/allocated 
[FG:InPlacePodVerticalScaling] Rework handling of allocated resources
 2024-10-25 23:24:52 +01:00
Tim Allclair
b186c160ca Clarify eviction based on allocated pods 2024-10-25 13:53:11 -07:00
Tim Allclair
c75a3e717e More precise allocatedPod name usage 2024-10-25 13:32:36 -07:00
Kubernetes Prow Robot
d3fd5940e4 Merge pull request #128333 from AnishShah/eviction-manager 
Revert #126562 that is causing eviction tests to fail
 2024-10-25 21:22:52 +01:00
Anish Shah
9faab07dda fix lint errors 2024-10-25 11:15:31 -07:00
Kubernetes Prow Robot
119f114f01 Merge pull request #128196 from richabanker/move-version 
Move k8s.io/apiserver/pkg/util/version to component-base
 2024-10-25 18:33:01 +01:00
Tim Allclair
40595bd94b Fix FakeStatusManager SetPodAllocation 2024-10-25 09:51:42 -07:00
Kubernetes Prow Robot
aec2ea1877 Merge pull request #124609 from AxeZhan/refac 
Move some helper functions from api/v1 to component-helpers
 2024-10-25 17:26:52 +01:00
Kubernetes Prow Robot
ed67bd77a8 Merge pull request #128336 from soltysh/cronjob_timestamp_stable 
Promote job creation timestamp annotation to stable
 2024-10-25 16:14:52 +01:00
Maciej Szulik
db94009981 Promote job creation timestamp annotation to stable 
Signed-off-by: Maciej Szulik <soltysh@gmail.com>
 2024-10-25 16:09:54 +02:00
Kubernetes Prow Robot
b95484e330 Merge pull request #128214 from googs1025/chore/printers/ut 
chore(printers): add unit test coverage for printers part
 2024-10-25 13:46:53 +01:00
Anish Shah
ac82e8a005 Revert "Fix issue in split filesystem where layers are stored on same disk but different paths" 
This reverts commit aea0b90652.
 2024-10-25 02:09:55 -07:00
AxeZhan
2ffb568540 rename functions 2024-10-25 12:53:24 +08:00
AxeZhan
9fe847482d move functions 2024-10-25 10:45:28 +08:00
Richa Banker
9274a584b8 Split k8s.io/component-base/registry and add into k8s.io/component-base/version and k8s.io/component-base/featuregate 2024-10-24 19:09:30 -07:00
Tim Allclair
7166169c82 Tidy up handlePodResize 2024-10-24 16:35:28 -07:00
Tim Allclair
34cf754fe9 Pass allocatedPods to canAdmitPod 2024-10-24 16:31:49 -07:00
Tim Allclair
d1f1bf200c Add more comments 2024-10-24 15:51:19 -07:00
Kubernetes Prow Robot
7b7a7968d4 Merge pull request #125314 from enj/enj/i/proto_for_core 
Use protobuf for core clients
 2024-10-24 18:20:54 +01:00
Tim Allclair
321eff34f7 Rework allocated resources handling 2024-10-24 09:27:40 -07:00
googs1025
e854086d74 chore(printers): add unit test coverage for printers part 2024-10-24 22:43:55 +08:00
Kubernetes Prow Robot
721d66780b Merge pull request #128305 from adrianmoisey/cidr_release_on_node_delete 
Ensure that a node's CIDR isn't released until the node is deleted
 2024-10-24 15:21:05 +01:00
Kubernetes Prow Robot
0a62f0fd7b Merge pull request #128139 from Jefftree/revert-allowservicelb 
Revert removal of feature AllowServiceLBStatusOnNonLB and LockToDefault first
 2024-10-24 15:20:54 +01:00
Adrian Moisey
4d2f3ed8e6 Ensure that a node's CIDR isn't released until the node is deleted 
Fixes https://github.com/kubernetes/kubernetes/issues/127792

Fixes bug where a node's PodCIDR was released when the node was given a
delete time stamp, but was hanging around due to a finalizer.
 2024-10-24 13:19:34 +02:00
Kubernetes Prow Robot
e526a27118 Merge pull request #116388 from mxpv/shutdown 
Clean/refactor node shutdown manager
 2024-10-24 08:34:53 +01:00
Kubernetes Prow Robot
aa8f2878a5 Merge pull request #117943 from lowang-bh/lessFunCall 
improve: reduce function calling number
 2024-10-24 04:52:52 +01:00
Maksym Pavlenko
449f86b0ba Refactor node shutdown manager 
Signed-off-by: Maksym Pavlenko <pavlenko.maksym@gmail.com>
 2024-10-23 17:36:22 -07:00
Kubernetes Prow Robot
5b61f60846 Merge pull request #128290 from zhifei92/fix-flake-test 
Add a thread-safe logBuffer for the watchdog unit test.
 2024-10-24 01:02:51 +01:00
Kubernetes Prow Robot
050a4a13b9 Merge pull request #128291 from alculquicondor/rm-alculquicondor-job 
Remove alculquicondor from job approvers
 2024-10-23 23:59:01 +01:00
Kubernetes Prow Robot
71093a09c1 Merge pull request #128244 from gnufied/fix-fsgroup-behaviour 
Apply fsGroup when accessMode is ReadWriteOncePod
 2024-10-23 23:58:52 +01:00
Kubernetes Prow Robot
c73aeaf5b5 Merge pull request #128234 from aroradaman/kube-proxy-multi-listen 
kube-proxy: use netutils.MultiListen for healthz and metrics server
 2024-10-23 21:00:59 +01:00
Kubernetes Prow Robot
d7e5ff87e0 Merge pull request #128083 from carlory/fix-126662-kubelet 
kubelet: fix a bug where kubelet wrongly drops the QOSClass field of the Pod's s status when it rejects a Pod
 2024-10-23 21:00:53 +01:00
Patrick Ohly
d53cb79cec DRA cel: enforce runtime limit by default again 
As pointed out during code review, the CEL cost estimates are not considered
perfectly reliable. Therefore it is better to also do runtime checks.

Some downstream users might decide to allow CEL expressions to run
longer. Therefore the cost limit is now part of an Options struct.
kube-scheduler uses the default cost limit defined in the resource.k8s.io API,
which is the same cost limit that also the apiserver uses during validation.
 2024-10-23 21:24:45 +02:00
Patrick Ohly
f548fc2264 DRA API: implement CEL cost limit 
The main purpose is to protect against denial-of-service attacks.  Scheduling
time depends a lot on unpredictable factors and expected scheduling time also
varies, so no attempt is made to limit the overall time spent on evaluating CEL
expressions per claim.
 2024-10-23 21:24:45 +02:00
Aldo Culquicondor
5fab6175b7 Remove alculquicondor from job approvers 
Change-Id: I2b1514ff70108602a589522cbb63dcdc88849313
 2024-10-23 17:58:55 +00:00
Kubernetes Prow Robot
352056f09d Merge pull request #127757 from torredil/scheduler-bugfix-5123 
scheduler: Improve CSILimits plugin accuracy by using VolumeAttachments
 2024-10-23 18:12:52 +01:00
carlory
c7e384f9ff kubelet: fix a bug where kubelet drops the QOSClass field of the Pod's status when it rejects a Pod 
Co-authored-by: Sergey Kanzhelev <S.Kanzhelev@live.com>
 2024-10-24 01:01:04 +08:00
Monis Khan
e4d10cfb1b Add test to confirm default content type used by core client 
Signed-off-by: Monis Khan <mok@microsoft.com>
 2024-10-23 11:35:32 -04:00
Monis Khan
6595fa4026 Fix tests that assume core clients use JSON 
Signed-off-by: Monis Khan <mok@microsoft.com>
 2024-10-23 11:35:30 -04:00
zhifei92
80a0096ee2 fix: Add a thread-safe logBuffer for the watchdog unit test. 2024-10-23 22:30:32 +08:00
Kubernetes Prow Robot
c6669ea7d6 Merge pull request #127155 from ffromani/alignment-metrics 
node: metrics: add resource alignment metrics
 2024-10-23 09:54:58 +01:00
Kubernetes Prow Robot
743f5a1a02 Merge pull request #127038 from oscgu/chore/rm-ioutil-3 
chore(client): rm ioutil
 2024-10-23 08:30:52 +01:00
Francesco Romani
c025861e0c node: metrics: add resource alignment metrics 
In order to improve the observability of the resource management
in kubelet, cpu allocation and NUMA alignment, we add more metrics
to report if resource alignment is in effect.

The more precise reporting would probably be using pod status,
but this would require more invasive and riskier changes,
and possibly extra interactions to the APIServer.

We start adding metrics to report if containers got their
compute resources aligned.
If metrics are growing, the assingment is working as expected;
If metrics stay consistent, perhaps at zero, no resource
alignment is done.

Extra fixes brought by this work
- retroactively add labels for existing tests
- running metrics test demands precision accounting to avoid flakes;
  ensure the node state is restored pristine between each test, to
  minimize the aforementioned risk of flakes.
- The test pod command line was wrong, with this the pod could not
  reach Running state. That gone unnoticed so far because
  no test using this utility function actually needed a pod
  in running state.

Signed-off-by: Francesco Romani <fromani@redhat.com>
 2024-10-23 08:05:38 +02:00
Kubernetes Prow Robot
cdc0cf5c10 Merge pull request #128237 from kolyshkin/userns 
pkg/kubelet/userns/inuserns: use moby/sys/userns
 2024-10-23 05:28:51 +01:00
Kubernetes Prow Robot
7fff5b6b02 Merge pull request #127566 from zhifei92/integrate-watchdog 
integrate kubelet with the systemd watchdog
 2024-10-23 02:21:34 +01:00
Kubernetes Prow Robot
0e6961e898 Merge pull request #126955 from tallclair/cleanup 
Remove corev1.Binding deprecation message
 2024-10-23 02:21:19 +01:00