Commit Graph

4577 Commits

Author SHA1 Message Date
Justin Santa Barbara 33b7432d6a Fix segfault when updating non-existent object
Fix #3935
2017-12-01 11:00:46 -05:00
Kubernetes Submit Queue 398c4ceebf
Merge pull request #3977 from justinsb/iptables_forward
Automatic merge from submit-queue.

Ensure iptables forwarding is enabled
2017-11-30 18:39:07 -08:00
Justin Santa Barbara 7fa4c28b1b Ensure iptables forwarding is enabled
Docker 1.13 changed how it set up iptables in a way that broke
forwarding.

We previously got away with it because we set the ip_forward sysctl,
which meant that docker wouldn't change the rule.  But if we're using an
image that preinstalled docker, docker might have already reconfigured
iptables before we run, and we didn't set it back.

We now set it back.

https://github.com/kubernetes/kubernetes/issues/40182
2017-11-30 20:29:32 -05:00
Kubernetes Submit Queue cf70f997e2
Merge pull request #3968 from justinsb/fix_typo_in_kops_get_cluster
Automatic merge from submit-queue.

Fix typo in kops get cluster --full hint
2017-11-30 15:13:11 -08:00
Kubernetes Submit Queue 0ef2fde69d
Merge pull request #3969 from justinsb/nvme_support_2
Automatic merge from submit-queue.

Initial support for nvme
2017-11-30 08:47:12 -08:00
Kubernetes Submit Queue fe976ac568
Merge pull request #3959 from so0k/addon-kubernetes-dashboard-fix
Automatic merge from submit-queue.

Add kubernetes-dashboard addon version constraint

This PR is to fix kubernetes-dashboard addon for kubernetes 1.7 clusters.

Before Patch, running `channels apply channel kubernetes-dashboard` for a k8s 1.7 cluster results in:

```bash
error validating data: unknown object type schema.GroupVersionKind{Group:"rbac.authorization.k8s.io", Version:"v1", Kind:"Role"}
``` 

After patch:
```
NAME                    CURRENT UPDATE
kubernetes-dashboard    -       1.7.1
I1129 06:41:10.314821   20137 addon.go:129] Applying update from "/home/training/kops/addons/kubernetes-dashboard/v1.7.1.yaml"
Updated "kubernetes-dashboard" to 1.7.1
```
2017-11-29 22:51:07 -08:00
Justin Santa Barbara 3006795aae Initial support for nvme 2017-11-30 00:23:43 -05:00
Justin Santa Barbara 2ada501c65 Fix typo in kops get cluster --full hint 2017-11-29 21:04:28 -05:00
Kubernetes Submit Queue fd7acd1ce6
Merge pull request #3948 from chrislovecnm/openapi-updates
Automatic merge from submit-queue.

openapi doc updates
2017-11-29 09:49:26 -08:00
Kubernetes Submit Queue 33136f99d6
Merge pull request #3960 from chrislovecnm/kops-branding
Automatic merge from submit-queue.

kops not Kops
2017-11-29 08:06:30 -08:00
chrislovecnm 3382b2d82f kops not Kops 2017-11-29 00:38:49 -07:00
Kubernetes Submit Queue 1bcf467bb3
Merge pull request #3949 from mikesplain/fix_brew_typo
Automatic merge from submit-queue.

Fix brew docs typo

Somehow the word kops got dropped when I modified these docs.
2017-11-28 23:31:18 -08:00
so0k a56b31aa51 Add kubernetes-dashboard addon version constraint 2017-11-29 14:44:16 +08:00
Kubernetes Submit Queue c98a771a97
Merge pull request #3956 from mikesplain/flannel_cpu_bug
Automatic merge from submit-queue.

Fix flannel error on starting

Fixes an issue where flannel does not startup due to limited resources:
```
kube-system   21m        21m         4         kube-flannel-ds                                       DaemonSet                                               Warning   FailedPlacement           daemonset-controller                      failed to place pod on "ip-10-25-40-53.ec2.internal": Node didn't have enough resource: cpu, requested: 104857600000, used: 0, capacity: 2000
```
2017-11-28 19:55:05 -08:00
Mike Splain 96aefac318 Fix flannel error on starting 2017-11-28 21:39:08 -05:00
Kubernetes Submit Queue d8df54b98b
Merge pull request #3953 from mikesplain/fix_flannel_version
Automatic merge from submit-queue.

Fix flannel version

Breaking out from https://github.com/kubernetes/kops/pull/3952 since this is needed for Kops 1.8
2017-11-28 14:47:10 -08:00
Mike Splain a2aa1da816 Fix flannel version 2017-11-28 16:26:46 -05:00
Kubernetes Submit Queue 5abeb5cd02
Merge pull request #3946 from justinsb/rename_dashboard_manifest
Automatic merge from submit-queue.

Rename dashboard manifest so it has extension
2017-11-28 11:42:19 -08:00
Mike Splain 71a32f2cf9 Fix brew docs typo 2017-11-28 13:41:43 -05:00
chrislovecnm da10075d5c openapi doc updates 2017-11-28 11:37:49 -07:00
Justin Santa Barbara c20975cc68 Rename dashboard manifest so it has extension 2017-11-28 13:30:32 -05:00
Justin Santa Barbara b8b7287e4a Bump versions to 1.8.0-beta.2 2017-11-28 13:27:58 -05:00
Kubernetes Submit Queue 305ffcc1af
Merge pull request #3936 from maciaszczykm/patch-1
Automatic merge from submit-queue.

Add Dashboard v1.8.0 deployment
2017-11-28 09:45:30 -08:00
Kubernetes Submit Queue be717cf0d2
Merge pull request #3940 from justinsb/bump_base_images
Automatic merge from submit-queue.

Bump all our base docker images
2017-11-28 08:30:27 -08:00
Marcin Maciaszczyk 11f4343407
Update v1.8.0 2017-11-28 10:29:52 +01:00
Marcin Maciaszczyk aca494822e
Update addon.yaml 2017-11-28 10:28:39 +01:00
Justin Santa Barbara 5a056a3872 Bump all our base docker images 2017-11-28 02:41:03 -05:00
Kubernetes Submit Queue 048a3ef770
Merge pull request #3939 from itskingori/add_hpa_use_clients_flag
Automatic merge from submit-queue.

Map horizontal-pod-autoscaler-use-rest-clients flag

Maps `--horizontal-pod-autoscaler-use-rest-clients` flag which is required for [Horizontal Pod Autoscaling][1]. See Kubernetes code in [kubernetes/kubernetes/blob/v1.7.11/cmd/kube-controller-manager/app/autoscaling.go#L36-L39][2].

Seems this is the missing piece for fulfilment of HPA pre-requisites, which are:

*  Enable the [Aggregation Layer][4] via the following kube-apiserver flags
   *  `--requestheader-client-ca-file=<path to aggregator CA cert>` (see https://github.com/kubernetes/kops/pull/3679)
   *  `--requestheader-allowed-names=aggregator` (see https://github.com/kubernetes/kops/pull/3679)
   *  `--requestheader-extra-headers-prefix=X-Remote-Extra-` (see https://github.com/kubernetes/kops/pull/3679)
   *  `--requestheader-group-headers=X-Remote-Group` (see https://github.com/kubernetes/kops/pull/3679)
   *  `--requestheader-username-headers=X-Remote-User` (see https://github.com/kubernetes/kops/pull/3679)
   *  `--proxy-client-cert-file=<path to aggregator proxy cert>` (see https://github.com/kubernetes/kops/pull/3165)
   *  `--proxy-client-key-file=<path to aggregator proxy key>` (see https://github.com/kubernetes/kops/pull/3165)
*  [Horizontal Pod Scaling][3] ... set the appropriate flags for `kube-controller-manager`:
  *   `--horizontal-pod-autoscaler-use-rest-clients` should be `true`.
  *  `--kubeconfig <path-to-kubeconfig>` (already set)

**Relevant Documentation:**

* https://v1-7.docs.kubernetes.io/docs/tasks/run-application/horizontal-pod-autoscale/
* https://v1-7.docs.kubernetes.io/docs/tasks/access-kubernetes-api/configure-aggregation-layer/

**Relevant Issues & PRs:**

* https://github.com/kubernetes/kops/pull/3679
* https://github.com/kubernetes/kops/issues/3152
* https://github.com/kubernetes/kops/issues/2691
* https://github.com/kubernetes/kops/issues/2652
* https://github.com/kubernetes/kops/pull/3165

[1]: https://v1-7.docs.kubernetes.io/docs/tasks/run-application/horizontal-pod-autoscale/
[2]: https://github.com/kubernetes/kubernetes/blob/v1.7.11/cmd/kube-controller-manager/app/autoscaling.go#L36-L39
[3]: https://v1-7.docs.kubernetes.io/docs/tasks/run-application/horizontal-pod-autoscale/
[4]: https://v1-7.docs.kubernetes.io/docs/tasks/access-kubernetes-api/configure-aggregation-layer/
2017-11-27 22:01:42 -08:00
Kubernetes Submit Queue 967ed4acf5
Merge pull request #3928 from justinsb/kopeio_auth
Automatic merge from submit-queue.

Update kopeio auth
2017-11-27 15:02:12 -08:00
King'ori Maina 28ff1bfe06
Add API machinery generated code 2017-11-27 23:12:33 +02:00
Kubernetes Submit Queue fe1ab92dac
Merge pull request #3932 from justinsb/bump_alpha
Automatic merge from submit-queue.

Bump alpha channel k8s versions
2017-11-27 12:03:45 -08:00
King'ori Maina 2d59315b96
Map horizontal-pod-autoscaler-use-rest-clients flag 2017-11-27 21:41:27 +02:00
Kubernetes Submit Queue 6a031a2b00
Merge pull request #3930 from justinsb/put_18_image_into_alpha
Automatic merge from submit-queue.

Put the 1.8 image into the alpha channel
2017-11-27 11:28:49 -08:00
Kubernetes Submit Queue 2967daabb2
Merge pull request #3931 from justinsb/promote_to_stable
Automatic merge from submit-queue.

Promote 1.5.8 and 1.6.11 to stable channel
2017-11-27 10:29:23 -08:00
Marcin Maciaszczyk 2971a60f6f
Add Dashboard v1.8.0 deployment 2017-11-27 13:11:01 +01:00
Justin Santa Barbara 31326059b1 Bump alpha channel k8s versions 2017-11-26 23:35:27 -05:00
Justin Santa Barbara 0112cc225e Promote 1.5.8 and 1.6.11 to stable channel 2017-11-26 23:31:40 -05:00
Justin Santa Barbara 660c45a01c Add initial docs on the kops side of authentication 2017-11-26 23:28:33 -05:00
Justin Santa Barbara 726ce3651f Put the 1.8 image into the alpha channel 2017-11-26 23:11:44 -05:00
Kubernetes Submit Queue 0a2f949fd9
Merge pull request #3929 from justinsb/add_service_extension_if_not_exists
Automatic merge from submit-queue.

Don't add .service extension if already there
2017-11-26 16:45:29 -08:00
Kubernetes Submit Queue e33a3ecee5
Merge pull request #3923 from justinsb/lockdown_etcd_peer_ports
Automatic merge from submit-queue.

Block etcd peer port from nodes
2017-11-26 16:17:13 -08:00
Kubernetes Submit Queue 8f0566d8a8
Merge pull request #3926 from justinsb/kubernetes_assets
Automatic merge from submit-queue.

Use EnsureTask so we don't have to track directories as closely
2017-11-26 15:43:20 -08:00
Kubernetes Submit Queue 0ade1dd2c4
Merge pull request #3822 from gambol99/dns_controller_optional
Automatic merge from submit-queue.

DNS Controller Optional

The current implementation enforces a dns-controller is running; given the user can switch the make the kube-apiserver server Internal and then reuse the dns for the masterInternalName; this effectlively removes the need to run the service (assuming your not using it for pods, node and service dns)

- adding a disableDnsController to the ExternalDNS spec provides a toggle on the addon (name is definitely up for debate)
- the default behaviour remains, the dns-controller is always pushed as an addon
2017-11-26 15:03:47 -08:00
Kubernetes Submit Queue 13451ad650
Merge pull request #3920 from aleerizw/fix/add_service_extension
Automatic merge from submit-queue.

Added .service to hooks unit files

Recent versions of systemd (version 229 at least) included in Ubuntu
16.04 and Debian 9 require the systemd unit files to have a .service
extension.

Fixes #3919

Signed-off-by: Ali Rizwan <ari@hellofresh.com>
2017-11-26 14:29:24 -08:00
Justin Santa Barbara 079464c223 Don't add .service extension if already there 2017-11-26 17:05:59 -05:00
Justin Santa Barbara d0944714f4 Update kopeio auth 2017-11-26 15:32:14 -05:00
Justin Santa Barbara b2cd5c961c Use EnsureTask so we don't have to track directories as closely
Issue #3921
2017-11-26 01:49:19 -05:00
Kubernetes Submit Queue b87689d844
Merge pull request #3925 from justinsb/avoid_ca_create_race
Automatic merge from submit-queue.

Avoid generating a CA keypair on-demand
2017-11-25 21:12:41 -08:00
Kubernetes Submit Queue 36abd6f418
Merge pull request #3922 from justinsb/bazel_fix_tests
Automatic merge from submit-queue.

bazel: fix tests/ directory
2017-11-25 20:41:24 -08:00
Justin Santa Barbara e3c7f03aaa Avoid generating a CA keypair on-demand
Instead we must explicitly create it; this avoids races where we are
reading the private key and creating CA certs.

Issue #3875
2017-11-25 23:12:55 -05:00