github/kubernetes
1
0
Fork 0
mirror of https://github.com/k3s-io/kubernetes.git synced 2025-08-08 03:33:56 +00:00
Code Issues Projects Releases Wiki Activity
126,828 Commits 42 Branches 7,905 Tags 1.3 GiB
7243fa60e1
Commit Graph

126828 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Kubernetes Prow Robot
5fcef4f79d
Merge pull request #128422 from bart0sh/PR163-density-e2e_node-adjust-limits 
density test: adjust CPU and memory limits
 2024-10-30 02:37:31 +00:00
Kubernetes Prow Robot
db66e397d9
Merge pull request #128359 from matteriben/disable-caching-for-authoritative-zone 
disable caching for authoritative zone to comply with rfc-1035 section 6.1.2
 2024-10-30 02:37:24 +00:00
Kubernetes Prow Robot
a93e3e7ae1
Merge pull request #127483 from nokia/strict-cpu-reservation-core 
KEP-4540: Add CPUManager policy option to restrict reservedSystemCPUs to system daemons and interrupt processing
 2024-10-30 01:21:47 +00:00
Kubernetes Prow Robot
d702d265c7
Merge pull request #127291 from zhifei92/fix-apiserver-unexpected-panic 
[FG:InPlacePodVerticalScaling] Fixed the apiserver panic issue that occurred when adding a container during pod updates in the InPlacePodVerticalScaling scenario.
 2024-10-30 01:21:40 +00:00
Kubernetes Prow Robot
a0e5e244b3
Merge pull request #126875 from serathius/watchcache-test-indexers 
Adding tests for using indexers in tests
 2024-10-30 01:21:32 +00:00
Kubernetes Prow Robot
6737352b03
Merge pull request #125708 from hshiina/dopodresizeaction-error 
[FG:InPlacePodVerticalScaling] Fix order of resizing pod cgroups in doPodResizeAction()
 2024-10-30 01:21:25 +00:00
Kubernetes Prow Robot
e8a75ac53f
Merge pull request #128420 from tallclair/e2e-cleanup 
Reuse cached client config for exec requests in e2e
 2024-10-30 00:17:37 +00:00
Kubernetes Prow Robot
42b7cfecec
Merge pull request #128274 from eddycharly/fix-cel-type-provider 
fix: cel type provider should return a type type
 2024-10-30 00:17:30 +00:00
Kubernetes Prow Robot
a339a36a36
Merge pull request #127506 from ffromani/cpu-pool-size-metrics 
node: metrics: add metrics about cpu pool sizes
 2024-10-30 00:17:24 +00:00
James Sturtevant
ac174f518c
Respond to sig-node feedback 
Signed-off-by: James Sturtevant <jsturtevant@gmail.com>
 2024-10-29 16:56:37 -07:00
Ed Bartosh
04f7a86001 density test: adjust CPU and memory limits 
Adjusted limits based on recent job log:
I1028 20:05:42.079182 1002 resource_usage_test.go:199] Resource usage:
  container cpu(cores) memory_working_set(MB) memory_rss(MB)
  "kubelet" 0.024      22.17                  14.20
  "runtime" 0.041      409.70                 84.21

  I1028 20:05:42.079274 1002 resource_usage_test.go:206] CPU usage of containers:
  container 50th% 90th% 95th% 99th% 100th%
  "/"       N/A   N/A   N/A   N/A   N/A
  "runtime" 0.014 0.834 0.834 0.834 1.083
  "kubelet" 0.023 0.093 0.093 0.093 0.164

Increasing 95th percentile for runtime CPU usage should also make
pull-kubernetes-node-kubelet-containerd-flaky less flaky.
 2024-10-30 00:48:56 +02:00
Kubernetes Prow Robot
f087575f21
Merge pull request #127226 from myeunee/cleanup 
Clean up unnecessary else block and redundant variable assignment
 2024-10-29 22:41:25 +00:00
Matt Riben
30d9ed7203
disable caching for authoritative zone 
Signed-off-by: Matt Riben <matt.riben@swirldslabs.com>
 2024-10-29 17:10:07 -05:00
myeunee
9cc65ce872 Restrict cz variable scope within else clause 2024-10-30 06:31:06 +09:00
Kubernetes Release Robot
f01e0d64db CHANGELOG: Update directory for v1.32.0-alpha.3 release 2024-10-29 20:17:52 +00:00
myeunee
2faaedbe39 Refactor error handling for configz initialization 
Improved code readability and limited variable scope as per reviewer's suggestion.
 2024-10-30 04:53:51 +09:00
Marek Siarkowicz
711772a1e1 Adding tests for using indexers in tests 2024-10-29 20:22:16 +01:00
Kubernetes Prow Robot
988769933e
Merge pull request #128307 from NoicFank/bugfix-scheduler-preemption 
bugfix(scheduler): preemption picks wrong victim node with higher priority pod on it
 2024-10-29 19:05:02 +00:00
Kubernetes Prow Robot
a12a32cd12
Merge pull request #127146 from bart0sh/PR156-DRA-Kubelet-latency 
Kubelet: add DRA latency metrics
 2024-10-29 19:04:55 +00:00
Tim Allclair
2407a49956 Reuse cached client config for exec requests in e2e 2024-10-29 10:00:11 -07:00
Kubernetes Prow Robot
c3980f601c
Merge pull request #128267 from benluddy/cbor-response-negotiation 
KEP-4222: Test response content negotiation for each CBOR enablement state.
 2024-10-29 16:48:55 +00:00
Yuki Iwai
eca7ee877a Self nominate tenzen-y as a reviewer for the Job controller 
Signed-off-by: Yuki Iwai <yuki.iwai.tz@gmail.com>
 2024-10-30 01:14:47 +09:00
Kubernetes Prow Robot
b8e20b74dd
Merge pull request #128382 from carlory/rm-vac 
remove unused vac code
 2024-10-29 15:25:05 +00:00
Kubernetes Prow Robot
c5ccf59974
Merge pull request #128379 from pohly/dra-owners-wg-label 
DRA: add wg/device-management label automatically
 2024-10-29 15:24:57 +00:00
Kubernetes Prow Robot
eb445ac66c
Merge pull request #128414 from soltysh/improve_error 
Provide link with e2e guidelines when verity-test-code.sh fails
 2024-10-29 14:21:06 +00:00
Kubernetes Prow Robot
c83250d104
Merge pull request #126754 from serathius/watchcache-btree 
Reimplement watch cache storage with btree
 2024-10-29 14:20:58 +00:00
Kubernetes Prow Robot
d09d98e07c
Merge pull request #128022 from googs1025/cleanup/ut/preemption 
chore(scheduler): add unit test for framework preemption part
 2024-10-29 13:16:55 +00:00
Talor Itzhak
d64f34eb2c memorymanager: areMemoryStatesEqual helper 
perform the memoryStates comparison in helper function

Signed-off-by: Talor Itzhak <titzhak@redhat.com>
 2024-10-29 14:22:04 +02:00
Marek Siarkowicz
50d2fab279 Implement btree based storage indexer 2024-10-29 13:13:21 +01:00
Maciej Szulik
97fcb05374
Provide link with e2e guidelines when verity-test-code.sh fails 
Signed-off-by: Maciej Szulik <soltysh@gmail.com>
 2024-10-29 13:07:05 +01:00
NoicFank
68f7a7c682 bugfix(scheduler): preemption picks wrong victim node with higher priority pod on it. 
Introducing pdb to preemption had disrupted the orderliness of pods in the victims,
which would leads picking wrong victim node with higher priority pod on it.
 2024-10-29 19:50:55 +08:00
Talor Itzhak
7476f46d71 memorymanager: fix checkpoint file comparison 
For a resource within a group, such as memory,
we should validate the total `Free` and total `Reserved` size of the expected `machineState` and state restored from checkpoint file after kubelet start.
If total `Free` and total `Reserved` are equal, the restored state is valid.

The old comparison however was done by reflection.

There're times when the memory accounting is equals
but the allocations across the NUMA nodes are varies.

In such cases we still need to consider the states as equals.

Signed-off-by: Talor Itzhak <titzhak@redhat.com>
 2024-10-29 12:10:27 +02:00
holder
6709317ae2 chore: optimize code logic 
(cherry picked from commit 91a9a195ac0fe0e31301dc60af0ea868fc4756ff)
 2024-10-29 12:08:28 +02:00
holder
6d7a1226d5 update the test case name 
(cherry picked from commit de033352079c7d87417f88f073d6b7891e51e590)
 2024-10-29 12:08:23 +02:00
holder
39726b119f fix: fix state validate error after memorymanager with static policy start 
(cherry picked from commit b91951f847d0b159c9d8ef32688cc96489ac1884)
 2024-10-29 12:08:16 +02:00
Patrick Ohly
4419568259 DRA: treat AdminAccess as a new feature gated field 
Using the "normal" logic for a feature gated field simplifies the
implementation of the feature gate.

There is one (entirely theoretic!) problem with updating from 1.31: if a claim
was allocated in 1.31 with admin access, the status field was not set because
it didn't exist yet. If a driver now follows the current definition of "unset =
off", then it will not grant admin access even though it should. This is
theoretic because drivers are starting to support admin access with 1.32, so
there shouldn't be any claim where this problem could occur.
 2024-10-29 10:22:31 +01:00
Patrick Ohly
9a7e4ccab2 DRA admin access: add feature gate 
The new DRAAdminAccess feature gate has the following effects:
- If disabled in the apiserver, the spec.devices.requests[*].adminAccess
  field gets cleared. Same in the status. In both cases the scenario
  that it was already set and a claim or claim template get updated
  is special: in those cases, the field is not cleared.

  Also, allocating a claim with admin access is allowed regardless of the
  feature gate and the field is not cleared. In practice, the scheduler
  will not do that.
- If disabled in the resource claim controller, creating ResourceClaims
  with the field set gets rejected. This prevents running workloads
  which depend on admin access.
- If disabled in the scheduler, claims with admin access don't get
  allocated. The effect is the same.

The alternative would have been to ignore the fields in claim controller and
scheduler. This is bad because a monitoring workload then runs, blocking
resources that probably were meant for production workloads.
 2024-10-29 09:50:11 +01:00
Patrick Ohly
f3fef01e79 DRA API: AdminAccess in DeviceRequestAllocationResult 
Drivers need to know that because admin access may also grant additional
permissions. The allocator needs to ignore such results when determining which
devices are considered as allocated.

In both cases it is conceptually cleaner to not rely on the content of the
ClaimSpec.
 2024-10-29 09:50:07 +01:00
Kubernetes Prow Robot
5f594f4215
Merge pull request #128401 from tenzen-y/use-same-receiver-name 
Job: Consistentely use the same reveiver name in the controller
 2024-10-29 08:16:55 +00:00
carlory
a2b1fc0f7a Move PodRejectionStatus into e2e/node from e2e/common/node 2024-10-29 15:55:58 +08:00
Kubernetes Prow Robot
66b3dc1a38
Merge pull request #128400 from tenzen-y/use-uid-typed-instead-of-string 
Job: Refactor uncountedTerminatedPods to avoid casting everywhere
 2024-10-29 06:32:56 +00:00
carlory
6da0175857 fix 2024-10-29 14:09:06 +08:00
yunwang
b00f221a85 fix InPlacePodVerticalScaling restore bug: the content wrote to and read from file pod_status are different 2024-10-29 13:39:58 +08:00
Yuki Iwai
d4959d8d29 Job: Consistentely use the same reveiver name in the controller 
Signed-off-by: Yuki Iwai <yuki.iwai.tz@gmail.com>
 2024-10-29 14:11:10 +09:00
Yuki Iwai
a23e7a42d3 Job: Refactor uncountedTerminatedPods to avoid casting everywhere 
Signed-off-by: Yuki Iwai <yuki.iwai.tz@gmail.com>
 2024-10-29 13:12:35 +09:00
jessestutler
f7003c76b4 Refactor the dynamicResources struct to DynamicResources 2024-10-29 11:44:42 +08:00
yunwang0911
114d4df4b4
Update pkg/kubelet/status/state/state_checkpoint.go 
Co-authored-by: Tim Allclair <timallclair@gmail.com>
 2024-10-29 11:39:54 +08:00
yunwang0911
8edc80c470
Update pkg/kubelet/status/state/state_checkpoint_test.go 
Co-authored-by: Tim Allclair <timallclair@gmail.com>
 2024-10-29 11:39:41 +08:00
yunwang0911
9d18e900c8
Update pkg/kubelet/status/state/checkpoint.go 
Co-authored-by: Tim Allclair <timallclair@gmail.com>
 2024-10-29 11:39:32 +08:00
yunwang0911
b67a5623b1
Update pkg/kubelet/status/state/checkpoint.go 
Co-authored-by: Tim Allclair <timallclair@gmail.com>
 2024-10-29 11:39:18 +08:00