Upcoming

Fixed a bug on Azure format-ephemeral.service where it was running before mnt-resource.mount.
coreos node list is now more verbose about each node.
coreos [etcdctl|fleetctl] is using only role=services nodes.
Using discovery again but without the hassle that we had before.
coreos kill enhanced by starting node's destruction at the same time
coreos bootstrap enhanced by pooling different node creations per account
Added host key to coreos node list
Cloudconfig changed to make docker.service wait for flanneld.service
coreos node exec has now a better arguments parser.
services role is now called boss
Added a command which keeps your cluster working and secure - coreos doctor
Added a firewall rules management per node to close a security issue to avoid connections to etcd server from the outside world.
Using a pre release of flannel instead because of NAT support.
Added a service that checks for private and public ips, if they arent present, it will reboot node.
Added an Environment Variable for allowing persistent config path setting.

0.1.3

Fix ephemeral disk detection, formatting and mounting
Added logic for execute remote commands with detection and uploading of local files to remote server before executing command. Detection is made by appending : before a local path.
Added same features present on accountGet, such as random get, on nodeGet

discarded discovery from the equation, this now handles cluster initialize
added .bootstap for initialize a cluster on multiple accounts
using group instead of image for selecting release group channel
we are now using etcd2 instead of etcd
after each .create we now wait until node is up by default for a better flow control when creating lots of nodes
added coreos etcdctl -- args compability
some changes on docs/ adding bootstrap as default method for creating a cluster
added role options so we can specify different kind of coreos cluster architectures
added new accountGet features such as random or iterative pick
fixed problems related to update channel group settings
enriched metadata content for better future cluster node filtering
added a doc which specifies how to create a scalable and production ready cluster cross locations

Added coreos fleetctl -- args compability
Added docs and changed examples/ folder to docs/
Added support for multiple nodes creation on CLI
Add API .kill and coreos kill --i-am=sure methods for easy testing
In case discovery is being generated while there are nodes on config, an error is thrown to avoid cluster division.
All ids are now uuid32 instead of Util.uniqId

npm name allocation