uncordon upgraded nodes stuck in cordoned state

[Master]Add ssh to checkout git repo with ssh key

[master]Fix auto upgrade

Problem:
An unavailable node is always reconciled to desired
state bypassing the usual cordon->upgrade->uncordon
path. If they become ready later on, node could have
already been upgraded before it's marked for upgrade
(NodeConditionUpgraded is never set). So the node
never gets uncordoned since we rely on node conditions.

[Validation] - Jenkins - override folder properties with job parameters

[master] add drain before delete to nodepool

Don't process objects if list is empty

Update rke go mod

Remove db password

Problem:
If nothing comes back from the list we don't want to process the app

Solution:
Exit and don't create the app so we don't spam etcd

[master] Use namespace from the object when looking for more

Problem:
When syncing helm configMaps and secrets there is a constant loop of
updates to apps causing large amounts of load

Solution:
When attempting to compare the object to other versions the namespace
being used was based off the release, this is not the namespace the
config object lives in so there would always be zero results making it
look like we were always working on the most recent object. Use the ns
from the config object to look for more.

Serve all index.html from CDN when running on master

Fix proxy URL path

Add additional parameter to pass in based on job changes

[master] Move document ending to if statment

Problem:
Currently, this will leave a --- at the end of the document which
indicates another document is coming. This causes the k8s python client
to break as it expects another document

Solution:
Move the if check for the next section so the --- is only added if it's
needed

Redirect /verify-auth(-azure)?/.* to /.*

[master] adding dnspolicy and tolerations to shell/helm

Problem: boot a cluster with only etcd and controlplane nodes, with no
worker coredns doesn't come up and the cluster-agent can not talk back
to the rancher server url and make the cluster accessible. additionally,
targeting for the shell fails due to taints.
Fix: Add dnspolicy default to the cluster-agent so when it boots it will
use the node's dns settings. Add tolerations to the shell/helmop for
controlplane nodes so the pods will assign and the tunnels will com up.

Updated CATTLE_MACHINE_VERSION to v0.15.0-rancher51

Add agent-install-script setting

Allow /meta/proxy to use any secret the user has GET access to

Add /cacerts route similar to k3s

Treat k3s derived distros as import so the SA account can be updated