It actually slows down things dramatically when used in combination
with Ansible.

Some systems (e.g. CentOS on Azure) have requiretty in sudoers which makes
pipelining fail.

In order to enable offline/intranet installation cases:
* Move DNS/resolvconf configuration to preinstall role. Remove
  skip_dnsmasq_k8s var as not needed anymore.

* Preconfigure DNS stack early, which may be the case when downloading
  artifacts from intranet repositories. Do not configure
  K8s DNS resolvers for hosts /etc/resolv.conf yet early (as they may be
  not existing).

* Reconfigure K8s DNS resolvers for hosts only after kubedns/dnsmasq
  was set up and before K8s apps to be created.

* Move docker install task to early stage as well and unbind it from the
  etcd role's specific install path. Fix external flannel dependency on
  docker role handlers. Also fix the docker restart handlers' steps
  ordering to match the expected sequence (the socket then the service).

* Add default resolver fact, which is
  the cloud provider specific and remove hardcoded GCE resolver.

* Reduce default ndots for hosts /etc/resolv.conf to 2. Multiple search
  domains combined with high ndots values lead to poor performance of
  DNS stack and make ansible workers to fail very often with the
  "Timeout (12s) waiting for privilege escalation prompt:" error.

* Update docs.

Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>

Version 1.9 has reduced verbosity for federation dns queries
which flood container logs.

Add tags to allow more granular tasks filtering.
Add generator script for MD formatted tags found.
Add docs for tags how-to.

Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>

This will allow to use '-e docker_version=1.12' in ansible playbook
execution. It's also backward-compatible and will work with floating
docker_version format in custom yaml files.

Closes #702

The variale etcd_access_addresses is used to determine
how to address communication from other roles to
the etcd cluster.

It was set to the address that ansible uses to
connect to instance ({{ item }})s and not the
the variable:
  ip_access
which had already been created and could already
be overridden through the access_ip variable.

This change allows ansible to connect to a machine using
a different address than the one used to access etcd.

Fixes: #669

Override GCE sysctl in /etc/sysctl.d/99-sysctl.conf instead of
the /etc/sysctl.d/11-gce-network-security.conf. It is recreated
by GCE, f.e. if gcloud CLI invokes some security related changes,
thus losing customizations we want to be persistent.

Update cloud providers firewall requirements in calico docs.

Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>

When running legacy calicoctl we do not specify calico hostname in
calico-node container thus we should not specify it in CNI config.

Also move 'legacy_calicoctl' set_fact task to the top.

kube-controller-manager's cluster signing cert and key points by default to not
existing `/etc/kubernetes/ca/ca.pem` and `/etc/kubernetes/ca/ca.key` [docs][1]

[1]: http://kubernetes.io/docs/admin/kube-controller-manager/#options

In new `calicoctl` version nodes peering with routers is broken.
We need to use predictable node names for calico-node and the
same names in calico `bgpPeer` resources and CNI.

New calicoctl does not support --detach=false option, so we should
use a recommended way to run calico-node service:
http://docs.projectcalico.org/v2.0/usage/configuration/as-service

Closes #674, #675

ipv4 forwarding gets broken when restarting networking, which
breaks all networking for all pods.

Fixes #655.

This is a teporary solution for long-polling idle connections to
apiserver. It will make Nginx not cut them for the duration of expected
timeout. It will also make Nginx extremely slow in realizing that there
is some issue with connectivity to apiserver as well, so it might not be
perfect permanent solution.

* Add an option to deploy K8s app to test e2e network connectivity
  and cluster DNS resolve via Kubedns for nethost/simple pods
  (defaults to false).
* Parametrize existing k8s apps templates with kube_namespace and
  kube_config_dir instead of hardcode.
* For CoreOS, ensure nameservers from inventory to be put in the
  first place to allow hostnet pods connectivity via short names
  or FQDN and hostnet agents to pass as well, if netchecker
  deployed.

Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>

* Add dns_replicas, dns_memory/cpu_limit/requests vars for
dns related apps.
* When kube_log_level=4, log dnsmasq queries as well.
* Add log level control for skydns (part of kubedns app).
* Add limits/requests vars for dnsmasq (part of kubedns app) and
  dnsmasq daemon set.
* Drop string defaults for kube_log_level as it is int and
  is defined in the global vars as well.
* Add docs

Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>

We no longer need to use v0.22.0 for calicoctl since Kargo has
support for new calicoctl CLI format.

Also fixing condition logic for calico pool task.

When download_run_once with download_localhost is used, docker is
expected to be running on the delegate localhost. That may be not
the case for a non localhost delegate, which is the kube-master
otherwise. Then the dnsmasq role, had it been invoked early before
deployment starts, would fail because of the missing docker dependency.

* Fix that dependency on docker and do not pre download dnsmasq image
  for the dnsmasq role, if download_localhost is disabled.
* Remove become: false for docker CLI invocation because that's not
  the common pattern to allow users access docker CLI w/o sudo.
* Fix opt bin path hack for localhost delegate to ignore errors when
  it fails with "sudo password required" otherwise.
* Describe download_run_once with download_localhost use case in docs
  as well.

Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>

Use cloud-init config to replace /etc/resolv.conf with the
content for kubelet to properly configure hostnet pods.

Do not use systemd-resolved yet, see
https://coreos.com/os/docs/latest/configuring-dns.html


"Only nss-aware applications can take advantage of the
systemd-resolved cache. Notably, this means that statically
linked Go programs and programs running within Docker/rkt
will use /etc/resolv.conf only, and will not use the
systemd-resolve cache."

Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>

W/o this patch, the "Download containers" task may be skipped
when running on the delegate node due to wrong "when" confition.
Then it fails to upload nginx image to the nodes as well.

Fix download nginx dependency so it always can be pushed to
nodes when download_run_once is enabled.

Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>

When setting permission for containers download/upload dir we're
using `ansible_ssh_user`. But if playbook is executed without
user being explicitly set `ansible_ssh_user` may be undefined.
In such situations dir ownership will default to `ansible_user_id`

Closes: #644

According to http://kubernetes.io/docs/user-guide/images/

 :
By default, the kubelet will try to pull each image from the
specified registry. However, if the imagePullPolicy property
of the container is set to IfNotPresent or Never, then a local\
image is used (preferentially or exclusively, respectively).

Use IfNotPresent value to allow images prepared by the download
role dependencies to be effectively used by kubelet without pull
errors resulting apps to stay blocked in PullBackOff/Error state
even when there are images on the localhost exist.

Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>

Pre download all required container images as roles' deps.
Drop unused flannel-server-helper images pre download.
Improve pods creation post-install test pre downloaded busybox.
Improve logs collection script with kubectl describe, fix sudo/etcd/weave
commands.

Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>

Fix unreliable waiting for the apiserver to become ready.
Remove logfile mount to align with the rest of static pods
and because containers shall write logs to stdout only.

Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>

in the etcd handler, the reload etcd action
was called after ansible waits for etcd to be
up, this means that the health checks which are
called immediately after fail (resulting in the etcd
role always failing and never finishing)

This patch changes the order to move the 'wait for etcd
up' resource after the 'reload etcd resource', ensuring that
the service is up before the health check is called.

* Add download_localhost for the download_run_once mode, which is
  use the ansible host (a travis node for CI case) to store and
  distribute containers across cluster nodes in inventory.
  Defaults to false.
* Rework download_run_once logic to fix idempotency of uploading
  containers.
* For Travis CI, enable docker images caching and run Travis
  workers with sudo enabled as a dependency
* For Travis CI, deploy with download_localhost and download_run_once
  enabled to shourten dev path drastically.
* Add compression for saved container images. Defaults to 'best'.

Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>
Co-authored-by: Aleksandr Didenko <adidenko@mirantis.com>

This patch introduce `calico_network_backend` global variable,
which allow to describe alternative network backend.
Default behavior is unchanged.

Add one more step (task) to containers download/upload sequence -
copy saved .tar containers to ansible host (delegate_to: localhost).

Then upload images to target nodes. It uses synchronize module so
if ansible host (localhost) is the same host as kube-master[0] then
new task causes no issues and the copy to localhost process is
basically skipped.