Commit Graph

6465 Commits

Author SHA1 Message Date
Kubernetes Prow Robot
fd40d68a39 Merge pull request #123238 from MirrorShih/master
Fix infinite loop and replace channel with queue
2024-04-18 14:30:07 -07:00
Kubernetes Prow Robot
745f56ea2b Merge pull request #124215 from tk3fftk/patch-2
fix: add delay to log message when enqueueing a job
2024-04-18 03:24:36 -07:00
Kubernetes Prow Robot
eb2a59e8d8 Merge pull request #124214 from Monokaix/dev
fix wrong comments of dra
2024-04-18 03:24:28 -07:00
Kubernetes Prow Robot
3188cbd1b3 Merge pull request #124113 from krzysdabro/nodeipam-pooling
nodeipam: poll nodes immediately
2024-04-18 03:23:33 -07:00
Kubernetes Prow Robot
d2ce87eb94 Merge pull request #123938 from pohly/dra-structured-parameters-tests
DRA: test for structured parameters
2024-04-18 02:10:08 -07:00
Filip Křepinský
85d55b6737 fix stateful set pod recreation and event spam (#123809)
* fix pods tracking and internal error checking in statefulset tests

* fix stateful set pod recreation and event spam

- do not emit events when pod reaches terminal phase
- do not try to recreate pod until the old pod has been removed from
  etcd storage

* fix conflict race in statefulset rest update

statefulset controller does less requests per sync now and thus can
reconcile status faster, thus resulting in a higher chance for conflicts
2024-04-18 01:03:46 -07:00
Kubernetes Prow Robot
854af6aba6 Merge pull request #122411 from huww98/lift-mountedByNode
ad controller: lift nodeAttachedTo.mountedByNode
2024-04-18 00:00:14 -07:00
Kubernetes Prow Robot
ef2c682635 Merge pull request #122082 from carlory/remove-keep-terminated-pod-volumes
keep-terminated-pod-volumes flag on kubelet is removed
2024-04-17 23:59:54 -07:00
huweiwen
3a71fe57f7 ad controller: lift nodeAttachedTo.mountedByNode
optimize adc.nodeUpdate(). Time complexity reduced from O(n) to O(1), where n is the number of nodes.

Data stored in nodeAttachedTo.mountedByNode is now at actualStateOfWorld.inUseVolumes.

This refactor also ensures that we can record the state update even if the volume is not present in ASW yet.

The added BenchmarkNodeUpdate result is reduced from 28076923 to 16030 ns/op.
The previous BenchmarkPopulateActualStateOfWorld result is also reduced from 13s to 8s.
2024-04-11 15:35:17 +08:00
Xuzheng Chang
3e08030d53 fix wrong comments of dra
Signed-off-by: Xuzheng Chang <changxuzheng@huawei.com>
2024-04-09 09:41:25 +08:00
Hiroki Takatsuka
84ffc2fc3d fix(job_controller): add delay duration to log message when enqueueing job 2024-04-07 16:42:48 +09:00
Krzysztof Dąbrowski
170d143b1a Revert changes in tests 2024-03-30 20:47:29 +01:00
Krzysztof Dąbrowski
5e94ffe90b nodeipam: poll nodes immediately 2024-03-29 20:56:08 +01:00
Patrick Ohly
4126e37f08 dra controller: unit tests 2024-03-22 10:03:22 +01:00
Kubernetes Prow Robot
611dbaa055 Merge pull request #122790 from carlory/fix-121696
Fix flaky test: Test_Run_OneVolumeDetachFailNodeWithReadWriteOnce
2024-03-10 19:23:40 -07:00
Kubernetes Prow Robot
d3d06c3c7e Merge pull request #123826 from tenzen-y/use-fake-client-job-unit
Job: Use the fake clock in TestTrackJobStatusAndRemoveFinalizers
2024-03-08 15:11:13 -08:00
Yuki Iwai
f2508df279 Job: Use the fake clock in TestTrackJobStatusAndRemoveFinalizers
Signed-off-by: Yuki Iwai <yuki.iwai.tz@gmail.com>
2024-03-09 06:09:05 +09:00
Nilekh Chaudhari
9161302e7f feat: implements svm controller
Signed-off-by: Nilekh Chaudhari <1626598+nilekhc@users.noreply.github.com>
2024-03-08 19:25:10 +00:00
Kubernetes Prow Robot
7ea3d0245a Merge pull request #123516 from pohly/dra-structured-parameters
DRA: structured parameters
2024-03-07 19:24:48 -08:00
Kubernetes Prow Robot
364ef335db Merge pull request #123412 from tenzen-y/add-new-jobsuccesspolicy-api
Job: Support for the SuccessPolicy
2024-03-07 14:49:20 -08:00
Patrick Ohly
3de376ecf6 dra controller: support structured parameters
When allocation was done by the scheduler, the controller needs to do the
deallocation because there is no control-plane controller which could react to
"DeallocationRequested".
2024-03-07 22:22:13 +01:00
Yuki Iwai
e216742672 Job: Support for the JobSuccessPolicy (alpha)
Signed-off-by: Yuki Iwai <yuki.iwai.tz@gmail.com>
2024-03-08 05:49:09 +09:00
Kubernetes Prow Robot
2ec63e0d28 Merge pull request #123482 from sanposhiho/hpa-containerresource-graduation
graduate HPAContainerMetrics to stable
2024-03-06 17:37:36 -08:00
cici37
de506ce7ac Promote ValidatingAdmissionPolicy to GA. 2024-03-05 16:00:21 -08:00
Kubernetes Prow Robot
180c6bdc25 Merge pull request #122056 from dhenkel92/keep-pdb-condition
keep existing PDB conditions when updating status
2024-03-05 10:40:42 -08:00
MirrorShih
62f89972b6 Fix infinite loop and replace channel with queue 2024-03-05 17:32:11 +00:00
Michał Woźniak
e568a77a93 Support for the Job managedBy field (alpha) (#123273)
* support for the managed-by label in Job

* Use managedBy field instead of managed-by label

* Additional review remarks

* Review remarks 2

* review remarks 3

* Skip cleanup of finalizers for job with custom managedBy

* Drop the performance optimization

* imrpove logs
2024-03-05 09:25:15 -08:00
Gaurav Ghildiyal
ec6fd2befa Add options construct to EndpointSlice NewReconciler for the new trafficDistributionEnabled field 2024-03-04 15:40:22 -08:00
Gaurav Ghildiyal
51a3fa2e6f Start reconciling on the new field 2024-03-03 22:46:03 -08:00
Kubernetes Prow Robot
df366107d1 Merge pull request #123529 from thockin/go-workspaces
Go workspaces for k/k and k/staging/*
2024-03-01 08:43:03 -08:00
carlory
b47c73ee26 keep-terminated-pod-volumes flag on kubelet is removed 2024-03-01 18:42:15 +08:00
Patrick Ohly
1d653e6185 test: use cancelation from ktesting
The return type of ktesting.NewTestContext is now a TContext. Code
which combined it WithCancel often didn't compile anymore (cannot overwrite
ktesting.TContext with context.Context). This is a good thing because all of
that code can be simplified to let ktesting handle the cancelation.
2024-03-01 07:51:22 +01:00
Tim Hockin
6dbc754ed6 Retool typecheck to be simpler
Instead of walking paths ourselves, just let Go's packages library do
it.  This is a slight CLI change - it wants "./foo" rather than "foo".

This also flagged a few things which seem to be legit failures.
2024-02-29 22:07:00 -08:00
Kubernetes Prow Robot
67a06c2056 Merge pull request #122293 from mengjiao-liu/controller-reconsider-log-verbosity
kube-controller-manager: readjust log verbosity
2024-02-29 11:55:21 -08:00
Kubernetes Prow Robot
12771ce773 Merge pull request #122493 from mengjiao-liu/fix-cronjob-error
cronjob: fix incorrect error checking
2024-02-27 14:14:54 -08:00
Michał Woźniak
3852d1c0c1 Make explicit check in CronJob if Job is successful before setting LastSuccessfulTime (#123380)
* Make explicit check in CronJob if Job is successful

before setting LastSuccessfulTime

* Review remarks for the CronJob

Co-authored-by: Filip Křepinský <fkrepins@redhat.com>

---------

Co-authored-by: Filip Křepinský <fkrepins@redhat.com>
2024-02-27 13:09:43 -08:00
Kubernetes Prow Robot
1853de77b2 Merge pull request #122874 from fusida/fix-kcm-panic
fix node lifecycle controller panic when conditionType ready is nil
2024-02-27 13:09:34 -08:00
Mengjiao Liu
b584b87a94 kube-controller-manager: readjust log verbosity
- Increase the global level for broadcaster's logging to 3 so that users can ignore event messages by lowering the logging level. It reduces information noise.
- Making sure the context is properly injected into the broadcaster, this will allow the -v flag value to be used also in that broadcaster, rather than the above global value.
- test: use cancellation from ktesting
- golangci-hints: checked error return value
2024-02-26 14:51:56 +08:00
fusida
9f6b48f1e7 fix node lifecycle controller panic when conditionType ready is nil 2024-02-26 11:26:45 +08:00
Kensei Nakada
07e0a80216 graduate HPAContainerMetrics to stable 2024-02-25 00:58:00 +00:00
Kubernetes Prow Robot
31a482a149 Merge pull request #120344 from rohitssingh/disable_force_detach
Add a flag to disable force detach behavior in kube-controller-manager
2024-02-22 13:02:38 -08:00
Rohit Singh
13dddca6a2 Add "disable-force-detach-on-timeout" flag to kube-controller-manager 2024-02-22 18:31:52 +00:00
Kubernetes Prow Robot
47f92ce231 Merge pull request #123242 from mimowo/fast-backoff-for-replacment-policy-tests
Improve accuracy of the PodsCreationTotal metric and use fast pod failure backoff for ReplacementPolicy integration tests
2024-02-15 09:54:53 -08:00
Michal Wozniak
f84d643c20 Use the Defer for pod replacement policy 2024-02-15 17:37:31 +01:00
Michal Wozniak
115dc90633 Increase accuracy of the pods_creation_total metric and improve test exec time 2024-02-15 10:59:01 +01:00
Kubernetes Prow Robot
68bbbd0359 Merge pull request #123300 from antoninbas/nodeipam-controller-remove-unused-field
Remove unused lookupIP field in NodeIPAM Controller struct
2024-02-14 17:00:13 -08:00
Kubernetes Prow Robot
684a9975fe Merge pull request #122919 from alexzielenski/apiserver/policy/mutating-initial
Refactor AdmissionPolicy for code sharing with mutating
2024-02-14 14:52:26 -08:00
Antonin Bas
75a3069294 Remove unused lookupIP field in NodeIPAM Controller struct
I am not sure why this was originally required, but it doesn't seem to
have been used for years.

Signed-off-by: Antonin Bas <antonin.bas@broadcom.com>
2024-02-14 13:28:35 -08:00
Kubernetes Prow Robot
bd0e1e2d9e Merge pull request #121554 from kaisoz/fix-job-controller-panics
Check if time is nil before logging
2024-02-13 09:33:04 -08:00
Alexander Zielenski
8b14116509 refactor: move vap into parent policy folder
also renames to remove stutter

comment
2024-02-12 10:58:24 -08:00