When your code changes, Smarx Role restarts the Ruby/Python/Node processes on each VM. Prior to this release, those processes were all restarted simultaneously, causing brief downtime for the app. With this release, one process is restarted at a time, waiting for up to a minute for the process to start responding to HTTP requests before moving on to the next process. This means that deployments configured to run more than one process (default is 4 in ServiceConfiguration.cscfg) shouldn't experience downtime during a code update.
To deploy, download the appropriate package based on the VM size you want, download and edit ServiceConfiguration.cscfg to specify the number of instances you want to run and the location of your application code, and deploy to Windows Azure.