hostweek.blogg.se - Kubeadm centos 7 vagrant

#Kubeadm centos 7 vagrant code#

it tries to use modern methods of deploying the "addons".kernel params like: -nf-call-iptables, etc.) plays nicely for corporate env: works via proxy.plays nicely for corporate env: allows use of internal registry for images (insted of using internet connection).KISS: it's build for kubeadm only (no other complexities arount it).creates Highly Available (HA cluster - multi master) (using VIPs) - using kubeadm.The project is for those who want to create&recreate k8s cluster using the official method (kubeadm), with all production features: Major difference from other projects: it uses kubeadm for all activities, and kubernetes is running in containers. Kubeadm simplifies drastically the installation, so for BYO (vms,desktops,baremetal), complex projects like kubespray/kops are not required any longer. Battle tested on for all Centos/RHEL 7.2+ till 7.6 and Ubuntu 16.04,18.04,19.10,20.04 (both with overlay2 and automatic docker_setup).Īctively used on a daily basis and tested with k8s starting 1.7 till 1.19. EFK - currently helm charts are not fully stable) Move from heapster to metrics server (once it will be stable).Authentication via LDAP (in plan Ke圜loak) integrate it in dashboard, grafana, etc.(more detailed comparison with other solutions towards the end of this readme)

any helm chart can be configured/added/removed via addons.yml.

actively tested on both Ubuntu/Debian and CentOS/RHEL.

enterprise-friendly: fully tested with http_proxy and private docker registry (usually private nexus registry proxy registry of docker.io, quay.io,, etc private mirror hostname&port fully configurable in this project).

Master HA & ingress setups accepts either: VIPs (using keepalived) or Hardware LB (when available).

start small with 1 vm, then add nodes, then make multi-master) -> all without downtime thanks to kubeadm.

scale UP or DOWN post deployment (e.g.

generates any cluster size, from 1 machine cluster (dev env) to productions sizes: all controlled by the provided inventory.

focused on "on-prem" deployments (but still accepts anything kubeadm can do) (vmware vsphere storage integration is actively used).

optionally, when docker_setup enabled, this project will also setup the docker with known kernel params for os (those from the k8s docs).

segregate nodes in 3 categories (when possible): masters, infra, compute (infra holds ingress controller, prometheus, grafana, and similar support tools) drives users towards good practices: e.g.this project does not hold any "custom" addon, everything that is installed is fetched directly their official repos (mostly helm repos).kubernetes cluster platform: not only k8s, but also the important addons.pure kubeadm based (all needless complexity removed) the stronger kubeadm will be, the smaller this project!.We felt that what was missing was getting things before and after the cluster installation, to get an initial (but reasonable) platform up.

Kubeadm is so strong now, that complex projects don't make sense. Going beyond minikube, making your own (usually on prem) k8s cluster (with the usuall addons installed) is still too hard or needlessly complex.

Its vision is to find and integrate the best tools out there (while using KISS principle). It tries to bring together most (if not all) the steps to get from a freshly installed linux to a working k8s cluster. It's gluing: pure kubeadm, offical helm charts for various addons, fine-tunings from docs and best practices.Īll based purely on kubeadm and official helm charts.

#Kubeadm centos 7 vagrant code#

Kubeadm-playboook ansible project's code is on Github Quick What is it:įor 3 years we keep on gathering best guidelines and growing this project for best kubernetes cluster installation + addons.