ba615ff94e
race condition in download role under vagrant
using a shared folder can cause race conditions for the download
role as it tries to download files on all the nodes to the same
shared path. This adds a flag to run the tasks in the download
role on just one node.
8 years ago
d8bebcd201
Fix issue with check_certs playbook
check_certs task "Check_certs | Set 'sync_certs' to true" was failing
due to the dict not existing, this sets defaults that allows the
correct behavior of the conditionals.
8 years ago
ae5ff890d4
fix flannel deployment, remove docker bridge before restarting
8 years ago
f949bfd46c
remove need for baking image to get writeable /opt/bin
8 years ago
66d9a6ebbc
updated to use handlers
8 years ago
9b8a757526
missed a name update :)
8 years ago
962155e463
updated names and removed checks for rhel, as we already know we have systemd inside that play
8 years ago
c90c981bb2
updated to support all OSes using systemd
8 years ago
608e7dfab2
upgrade k8s vers, and add a script for future upgrades
8 years ago
1884d89d3b
fixes the certs issue when masters or not in the kube-node group
8 years ago
9f8466a186
ensure ALL certs are synced between masters
8 years ago
743ad0eb5c
s/sync_certs/sync_tokens
8 years ago
5253b3ec13
ensure ca.pem makes it to multi-masters
8 years ago
ebf8231c9a
only run if proxy variables are added
8 years ago
adceaf60e1
support proxies with docker on rhel
8 years ago
5f2fa6d76f
revert .gitignore for secrets
8 years ago
bd064e8094
fix flannel's cross vm networking for vagrant
* set flannel backend type to `host-gw`
* set flannel interface to be eth1 ip
8 years ago
8f4e879ca7
Add native Vagrant support
This allows you to simply run `vagrant up` to get a 3 node HA cluster.
* Creates a dynamic inventory and uses the inventory/group_vars/all.yml
* commented lines in inventory.example so that ansible doesn't try to use it.
* added requirements.txt to give easy way to install ansible/ipaddr
* added gitignore files to stop attempts to save unwated files
* changed `Check if kube-system exists` to `failed_when: false` instead of
`ignore_errors`
8 years ago
4f627baf71
generate secrets on first master
8 years ago
180f2d1fde
Pull correct variable for etcd initial variable
This shouldn't use the `inventory_hostname` variable, as that will just yield the same variable, but rather use the `host` which we're looping over.
8 years ago
47982ea21c
Use ansible array format instead of dot-notation.
This fixes the ansible error ```'dict object' has no attribute
'ansible_default_ipv4'"}```. Closes #215
8 years ago
97de82bbcc
upgrade weave to v1.5.0 with cni
8 years ago
928bbeaf0f
upgrade calico v0.19.0, calico-cni v1.2.1
8 years ago
fb980e4542
missed tick in message
cosmetic
8 years ago
f12ad6a56f
consolidate EPEL change into a single task
as per @smana request
8 years ago
831a54e9b7
add code to detect if epel is already installed and skip it if it is
add option for generic
8 years ago
3cd89bed45
Kubernetes upgrade to 1.2.2
8 years ago
48a85ce8f8
use docker repository to install on CentOS
8 years ago
936927a54f
Fix docker install on rhel7
8 years ago
5c22133492
fix add nodes to the cluster
8 years ago
b03093be73
update kubectl bash completion
change hyperkube image repository
8 years ago
bc44d5deb3
upgrade to kubernetes v1.2.1
8 years ago
850b7466cd
remove deprecation warns and update doc
8 years ago
ab21f4d169
Define empty default objects for docker package management.
On CoreOS where there is no package management, perform zero-trip
loops instead of throwing an exception for iterating over a member
of an undefined variable.
8 years ago
64a39fdb86
Use var for bin dir instead of assuming /usr/local/bin
On CoreOS the binaries are not installed in /usr/local/bin.
8 years ago
7237a925eb
Add kubernetes.default.svc in certs dns
8 years ago
8cbdf73eba
Changed path to hosts ssl certs from /usr/share/ca-certificates to /etc/ssl/certs/ which fixes https problems in kube-controller-manager and kube-apiserver ( #189 ) caused by the lack of certificates on debian and redhat based systems.
8 years ago
624a964cda
Implemented Dynamic Provisioning of PersistentVolumes with cinder
When kubespray is deployed on OpenStack, the kube-controller-manager is now aware of the cluster and can create new cinder volumes automatically if the PersistentVolumeClaims are annotated accordingly.
Note that this is an alpha feature of kubernetes 1.2
8 years ago
a192111e6a
Reverted deletion of "changed_when: False" for Task "Install python-dnf for latest RedHat versions". Deleted "changed_when: False" on Task "Install latest version of python-apt for Debian distribs" to get notified when kubespray has installed a new package version.
8 years ago
4271dd6645
using apt module instead of command module to install python-apt
8 years ago
457ed11b49
fixed deprecation warnings regarding bare variables
8 years ago
9f8da6c225
Implemented cloud-provider integration for OpenStack.
Currently kubespray does not install kubernetes in a way that allows cinder volumes to be used. This commit provides the necessary cloud configuration file and configures kubelet and kube-apiserver to use it.
8 years ago
68fafd030d
choose between gce and aws cloud providers
8 years ago
e8aec5f4f0
Don't call the apiserver when the server is master only
8 years ago
c51ed4bbb7
use master election option instead of podmaster
8 years ago
ba4ad51c26
add aufs-tools package for debian
8 years ago
785b84fd43
Upgrade to docker 1.10.3
8 years ago
15ce66b2f5
Kubernetes 1.2.0
8 years ago
7e6d7caf4b
Dnsmasq runs on all nodes
8 years ago
6297e5ea93
Use dnsmasq inside pods
8 years ago
Paul Czarkowski
Smana
Spencer Smith
David Reuss
Christopher M Luciano
Rob Hirschfeld
Stig Telfer
ant31
teuto.net Netzdienste GmbH
Smaine Kahlouch
Antoine Legrand