0.1.4 • Published 9 years ago
coreos v0.1.4
coreos
the missing CoreOS cli for creating and managing clusters on different providers
Warning:
Currently we only support Microsoft Azure Cloud. Feel free to PR your own account provider class.
Cli
Installation
npm install --global coreos
Usage
Please check --help
with cli for further details.
coreos --help
Examples:
# Change to home dir so we can save config on "~/.coreos.json"
cd ~;
# Add an Azure Account
coreos account add \
--provider="Azure" \
--subscription="xxxx-xxxxx-xxxxxxx...." \
--pem="/your/path/to/cer.pem";
# Say to CoreOS instance we will use this on next commands
coreos account setCurrent \
--account="{account-id}"
# Now, lets say we want to add an azure-based node into our cluster
coreos node create \
--location="West US"
# for instance you may want to access to node shell
coreos node ssh --node="{nodeid}"
# Or execute a command with local files (example, submitting an unit)
coreos node exec -- fleetctl submit :/my/local/unit.service
# Its easy right? If you think so, star this project!
Node.js / io.js
Installation
npm install --save coreos
Usage
Please check API for further methods details.
Examples:
var CoreOS = require( 'coreos' );
var cos = new CoreOS({
// Don't read configuration file
loadConfigOnInit: false,
});
// Add an Azure Account
var azure = new CoreOS.Account.Provider.Azure(
{
subscription: 'xxxx-xxxxx-xxxxxxx....',
pem: fs.readFileSync( '/your/path/to/cer.pem' );
},
// Notice we are passing current CoreOS instance
cos
);
// Say to CoreOS instance we will use this
// on next commands
azure.setAsCurrent();
// Now, lets say we want to add an azure-based node
// into our cluster
cos.nodeCreate({
location: 'West US',
})
// As it returns a promise (hell yeah!!), we can then
// do something with node's info
.then(function ( node ) {
console.log( node );
});
// Its easy right? If you think so, star this project!
Motivation
findhit entered Microsoft BizSpark program recently and we ended up choosing CoreOS as our cloud cluster OS. We did a bunch of internal bash scripts for handling cluster scaling and its resources but it ended up too hard to maintain.
As CTO, I've decided to create a cli manager as our scaling endpoint.
Contributing
Feel free to contribute by creating Issues or Pull Requests.