* node: fix default kubelet/runtime cgroups when kube_reserved is false (default)
Commit 1c4db6132d introduced a notion of
kube_reserved. This introduced a breaking change defaulting to use
kube.slice for the container_manager and the kubelet as if kube_reserved
was always enabled whereas it is disabled by default.
This commit fixes this by bringing back system.slice whenever
kube_reserved is disabled.
Signed-off-by: Arthur Outhenin-Chalandre <arthur.outhenin-chalandre@proton.ch>
* inventory/sample: change false for kube_reserved as its the default
Changing the commented value in sample inventory to the actual default
value.
Signed-off-by: Arthur Outhenin-Chalandre <arthur.outhenin-chalandre@proton.ch>
---------
Signed-off-by: Arthur Outhenin-Chalandre <arthur.outhenin-chalandre@proton.ch>
* network_plugin/custom_cni: add CNI to apply provided manifests
Add a new simple custom_cni to install provided Kubernetes manifests.
This could be useful to use manifests directly provided by a CNI when
there are not support by Kubespray (i.e.: helm chart or any other manifests
generation method).
Co-authored-by: James Landrein <james.landrein@proton.ch>
Signed-off-by: Arthur Outhenin-Chalandre <arthur.outhenin-chalandre@proton.ch>
* network_plugin/custom_cni: add test with cilium
Co-authored-by: James Landrein <james.landrein@proton.ch>
Signed-off-by: Arthur Outhenin-Chalandre <arthur.outhenin-chalandre@proton.ch>
---------
Signed-off-by: Arthur Outhenin-Chalandre <arthur.outhenin-chalandre@proton.ch>
Co-authored-by: James Landrein <james.landrein@proton.ch>
This commit uses a kubeadm join config to pull down cert for etcd in
workers nodes (which is needed in some circumstances, for instance with
calico or cilium).
The previous way didn't allow us to pass certain parameters which was
typically given in the config in other kubeadm invokations in Kubespray.
This made kubeadm produced some errors for some edge cases.
For example, in our deployment we don't have a default route and even
though it's only to download the certificates, kubeadm produce an error
`unable to select an IP from default routes` (these command are kubeadm
controlplane command, so kubeadm does some additional checks). This is
fixed by specifying `advertiseAddress` within the kubeadm config.
Signed-off-by: Arthur Outhenin-Chalandre <arthur.outhenin-chalandre@proton.ch>
Signed-off-by: Arthur Outhenin-Chalandre <arthur.outhenin-chalandre@proton.ch>
* optimize cgroups settings for node reserved
* fix
* set cgroup slice for multi container engine
* set cgroup slice for crio
* add reserved cgroups variables to sample files
* Compatible with cgroup path for different container managers
* add cgroups doc
* fix markdown
* Fix inconsistent handling of admission plugin list
* Adjust hardening doc with the normalized admission plugin list
* Add pre-check for admission plugins format change
* Ignore checking admission plugins value when variable is not defined
* Support kubeadm patches in v1beta3
* Update kubeadm patches sample files in inventory
* Fix pre-commit syntax
* Set kubeadm_patches enabled to false in sample inventory
* Add the option to enable default Pod Security Configuration
Enable Pod Security in all namespaces by default with the option to
exempt some namespaces. Without the change only namespaces explicitly
configured will receive the admission plugin treatment.
* Fix the PR according to code review comments
* Revert the latest changes
- leave the empty file when kube_pod_security_use_default, but add comment explaining the empty file
- don't attempt magic at conditionally adding PodSecurity to kube_apiserver_admission_plugins_needs_configuration
The commit 1ce2f04 tried to merge multiple SUSE OS checks including
"openSUSE Leap" and "openSUSE Tumbleweed" into a single SUSE, but
that was a perfect change.
Then the commit c16efc9 tried to fix it for "openSUSE Leap", but it
didn't take care of "openSUSE Tumbleweed".
Then this adds "openSUSE Tumbleweed" to the OS check.
* feat: make kubernetes owner parametrized
* docs: update hardening guide with configuration for CIS 1.1.19
* fix: set etcd data directory permissions to be compliant to CIS 1.1.12
* extra admission controls now don't have a version in their file names
eventratelimit.v1beta2.yaml.j2 -> eventratelimit.yaml.j2
* cri_socket variable includes the unix:// prefix to be conformat with
upstream
* Fix: set fallback value of kubelet ip6 (#8858)
* Prune the spurious comma in the end of kubelet_address
- Update `roles/kubernetes/node/defaults/main.yml`
Co-authored-by: Cristian Calin <6627509+cristicalin@users.noreply.github.com>
* Fix: set fallback value of kubelet ip6 (#8858)
- Apply the lint: 132606368e
Co-authored-by: Cristian Calin <6627509+cristicalin@users.noreply.github.com>
This reverts commit e375678674.
The workaround of explicitly specifying root for the kubelet unit was
for pulling images from private registry. Kubernetes now have a
dedicated mechanism with imagePullSecret.
Current Kubespray supports the Kubernetes version 1.21 or upper with
`kube_version_min_required: v1.21.0`
Then kube_version v1.20- related code is not used at all.
This deletes those code for cleanup.
* [etcd] Add extra documentation for `etcd_memory_limit` and `etcd_quota_backend_bytes`
Signed-off-by: necatican <necaticanyildirim@gmail.com>
* [etcd] Add support for setting ETCD_MAX_REQUEST_BYTES
Signed-off-by: necatican <necaticanyildirim@gmail.com>
* Assert that IP range is enough for the nodes
Co-authored-by: Necatican Yıldırım <necaticanyildirim@gmail.com>
* Fixed whitespace
* Fixed errors
* Fixed errors
Co-authored-by: Necatican Yıldırım <necaticanyildirim@gmail.com>
aufs-tools was required for docker.io package originally,
but Kubespray installs docker-ce package instead today.
In addition, Ubuntu 20.04 doesn't provide aufs-tools as [1].
Then this removes aufs-tools from Ubuntu requirement.
[1]: https://bugs.launchpad.net/ubuntu/+source/aufs-tools/+bug/1947004
If running Kubespray on static IP environments, a task was failed like:
TASK [kubernetes/preinstall : Configure dhclient hooks for resolv.conf (RH-only)]
fatal: [ak8s2]: FAILED! => {
"changed": false, "checksum": "..",
"msg": "Destination directory /etc/dhcp/dhclient.d does not exist"}
This adds a check for dhclientconffile for running 0100-dhclient-hooks to
run the task only if dhcpclient is enabled.
* [calico] make vxlan encapsulation the default
* don't enable ipip encapsulation by default
* set calico_network_backend by default to vxlan
* update sample inventory and documentation
* [CI] pin default calico parameters for upgrade tests to ensure proper upgrade
* [CI] improve netchecker connectivity testing
* [CI] show logs for tests
* [calico] tweak task name
* [CI] Don't run the provisioner from vagrant since we run it in testcases_run.sh
* [CI] move kube-router tests to vagrant to avoid network connectivity issues during netchecker check
* service proxy mode still fails connectivity tests so keeping it manual mode
* [kube-router] account for containerd use-case
* Update ansible-lint to 5.4.0 (#8607)
It seems that the Rich version 11.0.0 has a breaking change.
So need to update ansible-lint to 5.3.2 or later.
* Fix for ansible-lint no-changed-when rule (#8607)
If kubelet is run with systemd (as it always is when using kubespray),
it starts in systemd's /system.slice/kubelet.service cgroup.
This commit prevents a creation and usage of a second unrelated cgroup.
* Use sysctl_file_path variable for all sysctl_file locations
* Add sysctl_file_path variable to kubespay-defaults
* Remove previously used sysctl file locations if present
* Use explicit filename in roles/kubernetes/node/defaults/main.yml
* Defaults: use explicit value
* Ensure entries for 1.23 are added for supported_versions vars
* cri-o: add support for kubernetes 1.23 but still use cri-o 1.22
* kubescheduler-config: diferentiate config versions based on kube_version
* containerd: change default resolvconf_mode to host_resolvconf
* Wait for kube-apiserver to come back after pod refresh
* Handle resolv.conf gracefully
* Retain currently configured DNS entries to ensure we don't break the resolvers
* Suse uses wickedd for network management so no dhcp hooks
* Molecule: increase ansible timeout
* CI: Increase ansible timeout to 120s for Packet jobs
* Improve control plane scale flow (#13)
* Added version 1.20.10 of K8s
* Setting first_kube_control_plane to a existing one
* Setting first_kube_control_plane to a existing one
* change first_kube_master for first_kube_control_plane
* Ansible-lint changes
* Containerd: download containerd from upstream instead of using distro specific packages
split runc download to separate role
make bootstrap-os role deploy container-selinux and seccomp libraries
clean up package manager provided containerd
move variables to docker role that are no longer common with containerd
* Containerd: make molecule testing more relevant
* replace ubuntu18 with ubuntu20
* add centos8 and debian11 to molecule tests
* run kubernetes/preinstall role to ensure relevancy
of test including dependency packages
* CI: adjust test scenarios for downloaded containerd
* Add Rocky as a known OS
* Make sure Rocky includes bootstrap-centos.yml
* Update docs with Rocky Linux
* Rocky Linux wireguard and EPEL
* Rocky Linux in the list of supported distributions
to remove deprecation warning:
> Flag --feature-gates has been deprecated, This parameter should be set via the config file specified by the Kubelet's --config flag.
Kubespray deployment failed when using containerd backend on nodes that apparmor was not installed or previously removed. This PR ensure apparmor is installed by adding it into required_pkgs var.
Kubespray deployment failed when using containerd backend on nodes that apparmor was not installed or previously removed. This PR ensure apparmor is installed by adding it into required_pkgs var.
The path of kubeconfig should be configurable, and its default value
is /etc/kubernetes/admin.conf. Most paths of the file are configurable
but some were not. This make those configurable.
* Calico: make calico_min_version check relevant
* Calico: only check currently installed version against the oldest supported version by the previous release
On Debian 11, `ipset` just recommend `iptables` so on the system that apt is configured with `APT::Install-Recommends "0";` iptables will not install automatically.
* Fedora and RHEL use etc_t and the convention is <type_name>_t
* Docs: specify all values for preinstall_selinux_state
* CI: Add Fedora 34 with SELinux in enforcing mode
* Ansible: move to Ansible 3.4.0 which uses ansible-base 2.10.10
* Docs: add a note about ansible upgrade post 2.9.x
* CI: ensure ansible is removed before ansible 3.x is installed to avoid pip failures
* Ansible: use newer ansible-lint
* Fix ansible-lint 5.0.11 found issues
* syntax issues
* risky-file-permissions
* var-naming
* role-name
* molecule tests
* Mitogen: use 0.3.0rc1 which adds support for ansible 2.10+
* Pin ansible-base to 2.10.11 to get package fix on RHEL8
* Calico: align manifests with upstream
* allow enabling typha prometheus metrics
* Calico: enable eBPF support
* manage the kubernetes-services-endpoint configmap
* Calico: document the use of eBPF dataplane
* Calico: improve checks before deployment
* enforce disabling kube-proxy when using eBPF dataplane
* ensure calico_version is supported
follow new naming conventions for gcr's coredns image.
starting from 1.21 kubeadm assumes it to be `coredns/coredns`:
this causes the kubeadm deployment being unable to pull image, beacuse `v`
was also added in image tag, until the role `kubernetes-apps` ovverides
it with the old name, which is only compatible with <=1.7.
Backward comptability with kubeadm <=1.20 is mantained checking
kubernetes version and falling back to old names (`coredns:1.xx`) when
the version is less than 1.21
* rename ansible groups to use _ instead of -
k8s-cluster -> k8s_cluster
k8s-node -> k8s_node
calico-rr -> calico_rr
no-floating -> no_floating
Note: kube-node,k8s-cluster groups in upgrade CI
need clean-up after v2.16 is tagged
* ensure old groups are mapped to the new ones
* AlmaLinux: ansible>2.9.19 is needed to know about AlmaLinux
* AlmaLinux: identify as a centos derrivative
* AlmaLinux: add AlmaLinux to checks for CentOS
* Use ansible_os_family to compare family and not distribution
* Add KubeSchedulerConfiguration for k8s 1.19 and up
With release of version 1.19.0 of kubernetes KubeSchedulerConfiguration
was graduated to beta. It allows to extend different stages of
scheduling with profiles. Such effect is achieved by using plugins and
extensions.
This patch adds KubeSchedulerConfiguration for versions 1.19 and later.
Configuration is set to k8s defaults or to kubespray vars. Moving those
defaults to new vars will be done in following patch.
Signed-off-by: Maciej Wereski <m.wereski@partner.samsung.com>
* KubeSchedulerConfiguration: add defaults
Signed-off-by: Maciej Wereski <m.wereski@partner.samsung.com>
Kay Yan
ERIK
Marijn van der Giesen
Jack
Arthur Outhenin-Chalandre
panguicai
Maxime Leroy
Mykola Ulianytskyi (Nikolay Ulyanitsky)
Bas
Haitian Chen
蒋航
Tristan
Florian Ruynat
Mohamed Zaian
Cyclinder
mKlaris
Shelming.Song
Kevin Huang
tu1h
C-Romeo
Lukas Najman
Jiffs Maverick
lijin-union
William Turner
Wouter Goedhart
Cristian Calin
Kenichi Omichi
Huang Chen-Yi
William Turner
Eugene Artemenko
Emin AKTAS
Zhong Jianxin
Kei Kori
cleverhu
Krystian Młynek
Alessio Greggi
Tomas Zvala
Ho Kim
Samuel Liu
h9-HSFRQDH
Citrullin
Calin Cristian Andrei
Thearas
Max Gautier
Necatican Yıldırım
emiran-orange
Robin Wallace
Andy
David Louks
Elif Akyıldırım
Victor Morales
Julien Le Fur
Mathieu Parent
Unai Arríen
onock
kakkotetsu
Takuya Murakami
Tom Janson
Mac Chaffee
Nicolas Goudry
Alex
Ilya Margolin
Michael Schmitz
华忠啊
cyril-corbon
Romain ALBON
singeleaf
Alvaro Campesino
Hanna Bledai
Lubos Mercl
EDGsheryl
Pasquale Toscano
Gheorghe Isak
Omar Aloraini
Iago Santos
rtsp
Marcos Lorenzo
Wang Zhen
Hari Hud
Daniil Muidinov
Smita Srivastava
cola-zero
spaced
Shinerrs
flix444
Kasakaze
Pavel Martynov
MRoci
Sergey
Etienne Champetier
Maciej Wereski