Commit Graph

96544 Commits

Author SHA1 Message Date
yue9944882
812f13fbec matches specific usernames instead of "*" 2020-11-23 15:12:26 +08:00
Hristo Venev
c8c81be8af range_allocator: Test (lack of) double counting 2020-11-22 23:09:09 -08:00
Hristo Venev
ee581278bd cidrset: Add test for double counting 2020-11-22 23:09:09 -08:00
Hristo Venev
4d28391c24 Fix double counting of IP addresses
The range allocator in pkg/controller/nodeipam/ipam/range_allocator.go
may call Occupy() on the same range twice:

1. Just before subscribing to the NodeInformer
2. From a callback given to the NodeInformer soon after registration
2020-11-22 23:09:09 -08:00
yue9944882
ee31c93748 ap&f e2e: eliminates client-side rate-limiting 2020-11-23 11:19:20 +08:00
DP19
9e4642211a add e2e test for Service ExternalIPs 2020-11-22 00:57:29 +01:00
Antonio Ojea
120472032c kube-proxy: treat ExternalIPs as ClusterIP
Currently kube-proxy treat ExternalIPs differently depending on:
- the traffic origin
- if the ExternalIP is present or not in the system.

It also depends on the CNI implementation to
discriminate between local and non-local traffic.

Since the ExternalIP belongs to a Service, we can avoid the roundtrip
of sending outside the traffic originated in the cluster.

Also, we leverage the new LocalTrafficDetector to detect the local
traffic and not rely on the CNI implementations for this.
2020-11-22 00:54:33 +01:00
rootlh
42c00bc523 fix bug: concurrent map writes error 2020-11-22 01:40:51 +08:00
Kubernetes Prow Robot
b2ecd1b3a3
Merge pull request #96716 from tosi3k/bump-npd-to-v0.8.5
Bump node-problem-detector to v0.8.5
2020-11-21 06:33:33 -08:00
Kubernetes Prow Robot
ece591f722
Merge pull request #96758 from msau42/revert-84206-refactor/remove-mount-volume-check-orphaned-pod-cleanup
Revert "check volume directories instead of mounts for cleanupOrphanedPodDirs"
2020-11-20 22:37:33 -08:00
Kubernetes Prow Robot
8c7cd8a8cc
Merge pull request #96553 from AlexeyPerevalov/FixesKubeletCrashEmptyTopology
Fixes sigfault in case of empty TopologyInfo
2020-11-20 16:03:33 -08:00
Kubernetes Prow Robot
8095565176
Merge pull request #96627 from aojea/hostport
fix e2e conformance test predicates conflict hostport
2020-11-20 15:03:34 -08:00
Basant Amarkhed
293d4b7c48 Avoiding double parsing of ip/cidr strings and logging bad ips/cidrs 2020-11-20 22:22:55 +00:00
Kubernetes Prow Robot
2ea6828ea0
Merge pull request #96720 from liggitt/throttled-logger
Deflake ThrottledLogger test
2020-11-20 13:56:22 -08:00
Maciej Szulik
32bdf7feaa
cli-runtime: expose option to set discovery burst 2020-11-20 20:20:01 +01:00
Antoni Zawodny
8f2dd3aaab Bump node-problem-detector to v0.8.5 2020-11-20 20:00:25 +01:00
Jan Chaloupka
d4e89547f0 kubectl/pkg/cmd tests: set explicit namespace
Otherwise, the namespace gets set to what's a current namespace
in provided kubeconfig. Which can be different from default.
2020-11-20 19:09:56 +01:00
Michelle Au
25edb8bc69
Revert "check volume directories instead of mounts for cleanupOrphanedPodDirs" 2020-11-20 09:06:09 -08:00
Jan Safranek
62ecc68183 Lower the frequency of volume plugin deprecation warning
The warning is always logged (klog.Warningf), so make sure it's logged only
once per process. We've seen clusters where the deprecation warning was
about 20% of total logs.
2020-11-20 14:22:28 +01:00
Kubernetes Prow Robot
7d72ddca6e
Merge pull request #96705 from andyzhangx/resize-fix
fix: resize Azure disk issue when it's in attached state
2020-11-19 21:36:51 -08:00
Kubernetes Prow Robot
06b0179895
Merge pull request #96662 from wojtek-t/fix_starting_rv_test
Fix TestStartingResourceVersion flakiness
2020-11-19 19:14:51 -08:00
Andrew Sy Kim
c753c48f30 core/v1: deprecate Service topologyKeys
Signed-off-by: Andrew Sy Kim <kim.andrewsy@gmail.com>
2020-11-19 21:09:48 -05:00
Kubernetes Prow Robot
264a5332f1
Merge pull request #96686 from mikedanese/ipt
gce: move iptables rule to mangle
2020-11-19 17:06:51 -08:00
Kubernetes Prow Robot
f35df2f841
Merge pull request #96727 from liggitt/revert-timeout
Revert timeout changes
2020-11-19 16:10:51 -08:00
Jordan Liggitt
afd92b3b3e Revert "plumb context with request deadline"
This reverts commit 83f869ee13.
2020-11-19 18:15:04 -05:00
Jordan Liggitt
3e21057070 Revert "use default value when the specified timeout is 0s"
This reverts commit 0090e27bd3.
2020-11-19 18:13:37 -05:00
Jordan Liggitt
63d2bf72fa Revert "add e2e tests for request timeout"
This reverts commit 2e6cb784d4.
2020-11-19 18:13:28 -05:00
Kubernetes Prow Robot
d0398c395a
Merge pull request #96185 from roycaihw/dedup-owner-references
Apiserver dedups owner references before creating/updating objects
2020-11-19 14:54:52 -08:00
Kubernetes Prow Robot
1ea29cec9c
Merge pull request #96487 from ii/promote-lifecycle-test-deployment
Promote AppsV1Deployment resource lifecycle test - +6 endpoint coverage
2020-11-19 13:02:51 -08:00
Jordan Liggitt
c84df5d66e Deflake ThrottledLogger test 2020-11-19 13:59:27 -05:00
Kubernetes Prow Robot
18099e1ef7
Merge pull request #96495 from andrewsykim/dockershim-exec-context
kubelet: dockershim ExecSync should return context.DeadlineExeceeded on timeout
2020-11-19 09:06:51 -08:00
andyzhangx
0d38026687 fix: resize Azure disk issue when it's in attached state
fix comments
2020-11-19 14:27:10 +00:00
lixiaobing1
8b351ac9d0 fix typos in type functionShapePoint 2020-11-19 20:22:09 +08:00
Hemant Kumar
1a323279a2 Address review comments 2020-11-19 06:30:44 -05:00
Damien Grisonnet
b525f9e0ed kube-aggregator: fix apiservice availability gauge
When an apiservice is deleted, its relative
aggregator_unavailable_apiservice metric remains with the value of the
last availability observed. Hence, if an apiservice is deleted while
being unavailable, the metric remains marked as unavailable.
This presents some problems when alerting on unavailable apiservices
as deleted apiservices might trigger the alert indefinitely.

To solve this issue, the aggregator_unavailable_apiservice metric should
only reflect the availability of existing apiservices.

This is achievable by using a custom Collector instead of a GaugeVec and
create throw-away metrics based on an apiservice lister output. With
this approach, on deletion, the apiservice will not be listed anymore,
resulting in its availability metric not being exposed.

Signed-off-by: Damien Grisonnet <dgrisonn@redhat.com>
2020-11-19 11:20:42 +01:00
wojtekt
37b0004978 Fix TestStartingResourceVersion flakiness 2020-11-19 07:56:17 +01:00
Haowei Cai
7a6afe7989 make sure managedFields are written 2020-11-18 19:21:25 -08:00
Hemant Kumar
b6ae233c40 Add docs about process of discovering disks from new nodes 2020-11-18 22:13:03 -05:00
Hemant Kumar
46a57b0f5c Fix unknown dangling volumes 2020-11-18 21:32:02 -05:00
Kubernetes Prow Robot
e64ebe0131
Merge pull request #96690 from jingxu97/nov/testwindows
Add linuxonly on one multivolume test
2020-11-18 16:24:50 -08:00
chenyw1990
a8add50ab6 don't add pod to podQueue when the NodeName of pod is not empty 2020-11-19 08:01:59 +08:00
Kubernetes Prow Robot
379ed6644d
Merge pull request #96484 from aojea/e2etest
add e2e test for dual-stack secondary service IPs
2020-11-18 15:28:51 -08:00
Haowei Cai
40a65577c7 generated 2020-11-18 12:48:26 -08:00
Haowei Cai
8bcf34a203 unit and integration tests
apiserver dedups and adds warning in CREATE/UPDATE/PATCH requests;
also handles duplication caused by mutating admission.
2020-11-18 12:46:20 -08:00
Kubernetes Prow Robot
1c49b4425b
Merge pull request #96646 from adtac/apfe2e-2
APF e2e tests: add request drown-out fairness test
2020-11-18 12:45:37 -08:00
jay vyas
0663e190ec Update to include windows / description of how NodeOSDistro is used in the e2es 2020-11-18 15:45:03 -05:00
Haowei Cai
ffc54ed1d2 apiserver dedups owner references and adds warning
for CREATE and UPDATE requests, we check duplication before managedFields
update, and after mutating admission; for PATCH requests, we check
duplication after mutating admission
2020-11-18 12:35:45 -08:00
Jing Xu
2a568c95dd Add linuxonly on one multivolume test
This test is not working for windows yet due to commands issued in pod
are not available for windows

Change-Id: Ia0b03afd6dfe0bbb1ab00dc821775450a7e8ce54
2020-11-18 11:58:12 -08:00
Kubernetes Prow Robot
b381baab66
Merge pull request #96681 from tkashem/request-timout-e2e
Use default value when the specified timeout for a request is 0s
2020-11-18 11:44:05 -08:00
Mike Danese
7fc57a207e gce: move iptables rule to mangle
This avoids a conflict with rules that calico installs. Also, acquire
the lock everywhere.
2020-11-18 11:28:03 -08:00