A Node.js dashboard framework.
This was written to display real-time company metrics using portable hardware and a couple large plasma displays. We originally tried using Dashing, which is nice for displaying many simple data points on one page, but we wanted to display complicated data on many pages. Another major difference is that Dashing uses Ruby, Sinatra and Batman.js (and Coffeescript), whereas Dashboard.js uses Node, Express, and a custom JavaScript framework (and no Coffeescript).
If you haven't installed Homebrew already, do that:
ruby -e "$(curl -fsSL https://raw.github.com/Homebrew/homebrew/go/install)"
If you haven't installed Node.js, Redis, Grunt, and Foreman already, do that:
brew install node redis
npm -g install grunt-cli foreman
Install Dashboard.js:
git clone git@github.com:intentmedia/dashboard.js.git
cd dashboard.js
# Install node dependencies
npm install
# Create the services directory and copy over the example services to get started:
cp -R app/example_services app/services
To start everything (including redis):
nf start
If you already have redis:
nf start web
Dashboard.js services are encapsulated in the sense that server-side code, client-side code, templates, CSS, and static assets all live in the same directory. The only required file is client.js
, but dashboards that have a server-side component require a server.js
file.
app/
services/
service_id/
public/ # static assets, available at /services/[service_id]/blah.jpg
client.js # all client-side service code
server.js # all server-side service code
[foo].css # bundled and loaded after the framework CSS
[template_id].html # hogan template, available to any service
A service involves both client-side and server-side polling. The server is continually updating its data, and the client is continually asking the server for the most recent copy.
The easiest way to understand how a service works is by looking at the example services included. There are a number of examples included that demonstrate the basic concepts and what's possible out of the box.
var redis = require('redis');
var redisClient = redis.createClient();
var SERVICE_KEY = 'DASHBOARD_COUNTER';
// Set to true to disable the service's server-side component.
exports.disabled = false;
// Jobs are simply functions that are executed at some date or time interval.
exports.jobs = [
{
cron: '1 0 0 * * *', // every night at midnight EST
callback: resetCounter,
timeZone: 'America/New_York'
},
{
cron: '* * * * * *', // every second
callback: updateCounter
}
];
// args is an object passed to the server via the `requestData` parameter.
// responseToClient sends data back to the client which gets passed
// to the `render` function.
exports.serve = function (args, responseToClient) {
redisClient.get(SERVICE_KEY, function (err, data) {
responseToClient(data ? data : 0);
});
};
function resetCounter() {
redisClient.set(SERVICE_KEY, 0);
}
function updateCounter() {
updateKey(SERVICE_KEY, function (count) {
count = parseFloat(count) || 0;
count += (Math.random() * 10);
return count;
});
}
function updateKey(key, updateFn) {
redisClient.get(key, function (err, data) {
data = updateFn(data);
redisClient.set(key, data);
});
}
There is one client-side function used to register and configure a service.
Dashboard.Service.add({
// The unique id of this service.
id: 'Counter',
// The name in the service selector.
name: 'Counter',
// Disabled services don't show up in the service selector.
disabled: false,
// The url to ask for data. The URL should map to the service name (i.e. app/services/counter).
url: '/service/counter',
// How often to poll the server for new data, in milliseconds.
interval: 1000,
// Data to be sent to the server. Can be an object or a function.
requestData: function () { },
// Function called once, during service init.
init: function () { },
// Function called before a different service is initialized.
tearDown: function () { },
// Function called when we get data from the server.
render: function (data) {
// Assumes a my_template.html file exists in this or another service directory.
Dashboard.View.render('my_template', data);
},
// Options that can be used to interact with the rendered data.
// Use Dashboard.Service.option(service_id, option_key) to get the selected value.
options: {
lookback: {
name: 'Granularity',
control: 'slider',
value: 15,
min: 1,
max: 30
},
scaleType: {
name: 'Scale Type',
control: 'dropdown',
value: 'linear',
values: [
{ name: 'Linear', value: 'linear' },
{ name: 'Square Root', value: 'sqrt' }
]
},
filter: {
name: 'Filter',
control: 'filter',
key: 'name'
}
}
});
This is beta software, use at your own risk. If you have any issues or feature requests, we would love to know, please open a GitHub issue. Contributions and pull requests are also very welcome - we would especially like to include more out-of-the-box example services.
Dashboard.js was created by Andrew Childs.