github/kubernetes
1
0
Fork 0
mirror of https://github.com/k3s-io/kubernetes.git synced 2026-08-08 23:37:11 +00:00
Code Issues Projects Releases Wiki Activity
131,982 Commits 42 Branches 9,895 Tags
13b8daa06e97e44d5566464a2d7beb6322d4540f
Commit Graph

131982 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Kubernetes Prow Robot
13b8daa06e Merge pull request #133934 from pohly/automated-cherry-pick-of-#133926-origin-release-1.34 
Automated cherry pick of #133926: DRA kubelet: avoid deadlock when gRPC connection to driver goes idle
 2025-09-26 01:34:28 -07:00
Kubernetes Prow Robot
3b28d3ba09 Merge pull request #133905 from huww98/automated-cherry-pick-of-#133890-origin-release-1.34 
Automated cherry pick of #133890: kubelet/metrics: fix multiple Register call
 2025-09-26 01:34:19 -07:00
Kubernetes Prow Robot
7ad368caa6 Merge pull request #134270 from neolit123/automated-cherry-pick-of-#134265-origin-release-1.34 
Automated cherry pick of #134265: kubeadm: ensure waiting for apiserver uses a local client
 2025-09-26 00:22:19 -07:00
Lubomir I. Ivanov
e9f5e892f7 kubeadm: use JoinHostPort in WaitControlPlaneClient 
Using naive string concat with ":" does not work for IPv6.
Apply the fix to WaitForKubelet as well.
 2025-09-25 21:26:38 +02:00
Lubomir I. Ivanov
663664c617 kubeadm: ensure waiting for apiserver uses a local client 
When waiting for the kube-apiserver to report 'ok'
in the 'init' and 'join' phase 'wait-control-plane', a client
constructed from the 'admin.conf' is used. In the case of the
kube-apiserver, the discovery client is used so that
anonymous-auth works. But if 'admin.conf' is used as is,
it would point to the CPE and not the LAE.

Implement a new method WaitControlPlaneClient() for both
init.go and join.go that patches the 'Server' field in the
loaded v1.Config to point to the LAE, before constructing
a client set and using it in the kube-apiserver waiter.
 2025-09-25 15:57:55 +02:00
Kubernetes Prow Robot
b340e9f8a1 Merge pull request #134222 from cpanato/update-go-rel-134 
[release-1.34][go] Bump images, dependencies and versions to go 1.24.7 and distroless iptables
 2025-09-24 09:08:17 -07:00
Kubernetes Prow Robot
63a33f1a6f Merge pull request #134150 from sats-23/automated-cherry-pick-of-#133398-origin-release-1.34 
Automated cherry pick of #133398: Split preemption integration tests to mitigate the timeout
 2025-09-23 23:56:13 -07:00
Carlos Panato
09ddc5ebbf Bump images, dependencies and versions to go 1.24.7 and distroless iptables 
Signed-off-by: Carlos Panato <ctadeu@gmail.com>
 2025-09-23 15:23:17 +02:00
Maciej Skoczeń
b245b551f7 Split preemption integration tests to mitigate the timeout 2025-09-20 17:05:18 +05:30
Kubernetes Prow Robot
b74e83ecbe Merge pull request #134118 from jack4it/automated-cherry-pick-of-#134024-upstream-release-1.34 
Automated cherry pick of #134024: fix: use iifname for input interface name matches
 2025-09-19 12:56:14 -07:00
Jack Ma
07ac83c146 fix: use iifname for input interface name matches 
add tests to cover nftables; also fix NewDetectLocalByBridgeInterface
 2025-09-17 09:59:14 -07:00
Kubernetes Release Robot
16e3f7c179 Update CHANGELOG/CHANGELOG-1.34.md for v1.34.1 2025-09-09 20:00:28 +00:00
Kubernetes Release Robot
93248f9ae0 Release commit for Kubernetes v1.34.1 v1.34.1 2025-09-09 19:37:19 +00:00
Patrick Ohly
b97399727e DRA kubelet: avoid deadlock when gRPC connection to driver goes idle 
When gRPC notifies the kubelet that a connection ended, the kubelet tries to
reconnect because it needs to know when a DRA driver comes back. The same code
gets called when a connection goes idle, by default after 30 minutes. In that
and only that case the conn.Connect call deadlocks while calling into the gRPC
idle manager.

This can be reproduced with a new unit test which artificially shortens the
idle timeout. This fix is to move the Connect call into a goroutine because
then both HandleConn and Connect can proceed. It's sufficient that Connect
finishes at some point, it doesn't need to be immediately.
 2025-09-08 20:36:06 +02:00
Kubernetes Prow Robot
45a8851a7c Merge pull request #133901 from yongruilin/automated-cherry-pick-of-#133896-upstream-release-1.34 
Automated cherry pick of #133896: fix: Only warn for unrecognized formats on type=string
 2025-09-05 14:33:26 -07:00
胡玮文
76ab5f0664 kubelet/metrics: fix multiple Register call 
DRA also calls Register at pkg/kubelet/cm/container_manager_linux.go NewContainerManager(), causing volume stats collector being ignored.
Fix this by moving it out of `sync.Once()`, allowing multiple calls to `Register()` func.
 2025-09-05 14:54:05 +08:00
胡玮文
04fd1ada6f kubelet/metrics: misc optimization 2025-09-05 14:54:05 +08:00
yongruilin
7bd2900b1e fix: Only warn for unrecognized formats on type=string 2025-09-05 03:53:40 +00:00
Kubernetes Prow Robot
8c0988abb6 Merge pull request #133745 from jsafrane/automated-cherry-pick-of-#133425-release-1.34 
Automated cherry pick of #133425: Fix SELinux label comparison
 2025-09-04 08:31:17 -07:00
Kubernetes Prow Robot
4fa47837fb Merge pull request #133749 from xigang/automated-cherry-pick-of-#132477-upstream-release-1.34 
Automated cherry pick of #132477: Fix DaemonSet misscheduled status not updating on node taint changes
 2025-09-04 05:45:16 -07:00
Kubernetes Prow Robot
c45c5b3c8c Merge pull request #133783 from vpnachev/automated-cherry-pick-of-#133771-origin-release-1.34 
Automated cherry pick of #133771: Fix completion of resource names
 2025-09-04 03:37:14 -07:00
Kubernetes Prow Robot
48666c4adf Merge pull request #133879 from serathius/automated-cherry-pick-of-#133873-origin-release-1.34 
Automated cherry pick of #133873: Disable estimating resource size for resources with watch cache disabled
 2025-09-04 01:13:16 -07:00
Marek Siarkowicz
f779cf6381 Disable estimating resource size for resources with watch cache disabled 
Listing all keys from etcd turned out to be too expensive, negativly
impacting events POST latency. Events resource is the only resource that
by default has watch cache disabled and which includes very
large number of small objects making it very costly to list keys.

Expected impact:
* No apiserver_resource_size_estimate_bytes metric for events.
* APF overestimating LIST request cost to events. Fallback assumes
  object size of 1.5MB, meaning LIST events will always get maxSeats
 2025-09-03 23:55:22 +02:00
Kubernetes Prow Robot
cb97641ddb Merge pull request #133866 from serathius/automated-cherry-pick-of-#133817-origin-release-1.34 
Automated cherry pick of #133817: Disable collecting stats for resources not setting prefix to prevent error logs
 2025-09-03 12:07:16 -07:00
Marek Siarkowicz
96739661c7 Disable collecting stats for resources not setting prefix to prevent error logs 2025-09-03 09:23:47 +02:00
Kubernetes Prow Robot
674cbc8c45 Merge pull request #133753 from HirazawaUi/automated-cherry-pick-of-#133513-upstream-release-1.34 
Automated cherry pick of #133513: Add missing conversion for timeoutForControlPlane
 2025-09-02 23:33:15 -07:00
Marc Khouzam
2631a5f21e Fix linter 
Signed-off-by: Marc Khouzam <marc.khouzam@gmail.com>
 2025-08-29 15:37:14 +03:00
Marc Khouzam
94c0e47882 Fix completion of resource names 
The output format is now used by the `Complete()` function, so it must
be set before invoking said function.

The commit also adds a unit tests for this scenario.

Signed-off-by: Marc Khouzam <marc.khouzam@gmail.com>
 2025-08-29 15:37:14 +03:00
tom1299
08515c751b Fix missing control plane health check timeout 2025-08-28 22:56:12 +08:00
xigang
a75938d9e5 Fix DaemonSet misscheduled status not updating on node taint changes 
Signed-off-by: xigang <wangxigang2014@gmail.com>
 2025-08-28 18:07:40 +08:00
Jan Safranek
3ebbe57311 Add a note about Conflicts return value 2025-08-28 10:29:24 +02:00
Jan Safranek
98dca817c9 Fix SELinux label comparison 
The comparison of SELinux labels in KCM tolerates missing fields - the
operating system is going to default them from its defaults, but in KCM we
don't know what the defaults are.

But the OS won't default the last component, "level", which includes also
categories. Make sure that labels with a level set conflicts with level "",
that's what will conflict on the OS too.
 2025-08-28 10:29:24 +02:00
Jan Safranek
3e75f38819 Add missing cases to SELinuxMount tests 2025-08-28 10:29:24 +02:00
Kubernetes Release Robot
d133742146 Update CHANGELOG/CHANGELOG-1.34.md for v1.34.0 2025-08-27 10:36:11 +00:00
Kubernetes Release Robot
f28b4c9efb Release commit for Kubernetes v1.34.0 v1.34.0 2025-08-27 10:09:04 +00:00
Kubernetes Release Robot
a99d6bb777 Merge remote-tracking branch 'origin/master' into release-1.34 2025-08-26 20:11:19 +00:00
Kubernetes Prow Robot
28c8381aa5 Merge pull request #133695 from prezha/kube-proxy-iptables-startup-errors 
Fix kube-proxy iptables startup: log but don't exit if ipv4 or ipv6 is not available
 2025-08-26 09:22:10 -07:00
Predrag Rogic
f2170bc6c3 kube-proxy: log but don't exit if ipv4 or ipv6 is not available 2025-08-26 16:11:12 +01:00
Kubernetes Release Robot
38a66ad237 Merge remote-tracking branch 'origin/master' into release-1.34 2025-08-21 14:01:20 +00:00
Kubernetes Prow Robot
091f87c10b Merge pull request #133625 from danwinship/servicecidr-conformance 
Remove patch/update from ServiceCIDR API conformance test
 2025-08-21 03:31:05 -07:00
Kubernetes Release Robot
91e472f9fd Merge remote-tracking branch 'origin/master' into release-1.34 2025-08-21 02:00:49 +00:00
Kubernetes Release Robot
f47e9696d7 CHANGELOG: Update directory for v1.34.0-rc.2 release 2025-08-20 17:40:32 +00:00
Kubernetes Release Robot
78405b2c89 Update CHANGELOG/CHANGELOG-1.34.md for v1.34.0-rc.2 2025-08-20 17:40:32 +00:00
Kubernetes Release Robot
ed3ccdbe32 Release commit for Kubernetes v1.34.0-rc.2 v1.34.0-rc.2 2025-08-20 17:17:54 +00:00
Dan Winship
db01f94032 Remove patch/update from ServiceCIDR API conformance test 
They were already listed in ineligible_endpoints.yaml, so we shouldn't
be testing them here anyway.
 2025-08-20 09:19:55 -04:00
Kubernetes Release Robot
343b53a054 Merge remote-tracking branch 'origin/master' into release-1.34 2025-08-19 20:00:00 +00:00
Kubernetes Prow Robot
4e8b192b66 Merge pull request #133604 from serathius/watchcache-count 
Fix storage counting all objects instead of objects for resource
 2025-08-19 09:23:36 -07:00
Marek Siarkowicz
d49e3d67cb Fix storage counting all objects instead just objects for resource 2025-08-19 16:35:34 +02:00
Kubernetes Release Robot
4feac134a9 Merge remote-tracking branch 'origin/master' into release-1.34 2025-08-19 14:03:18 +00:00
Kubernetes Prow Robot
5568780ca3 Merge pull request #133562 from nojnhuh/dra-e2e-slice-controller-flake 
DRA: wait for stats to converge in "creates slices" e2e test
 2025-08-19 04:21:36 -07:00