Puppet supported by CloudML

[Stepan Seycek]

Dear Puppet users,

for those of you using Puppet for configuration management on top of IaaS the extension of the open source
multi-cloud modelling and an run-time tool CloudML may be of interest.

CloudML lets you define a deployment model for your applications consisting of

• cloud resources such as virtual machines
• external components such as PaaS services
• internal components which are the deployable artefacts
  
• and dependencies

Such a deployment model can be passed to the run-time tools as a JSON document, and the actual provisioning
of cloud resources and the deployment and configuration of the internal components can be triggered then. The
run-time tools also support continuous updates to the deployment model, where only the differences between two
versions are enacted if the provisioning process is triggered.

The modelling language and the deployment mechanisms have been extended with Puppet support recently. This
feature can be used to annotate the internal components in the deployment mode with Puppet manifest declarations
that are then being used for the creation of node manifests associated with the virtual machines.

If you are interested in further details, you can find information about CloudML at http://www.cloudml.org and about
its Puppet feature in particular at https://github.com/boc-modaclouds/cloudml-with-puppet. A major part of this work
has been done in the context of the EU co-founded research project MODAClouds (http://multiclouddevops.com/).

Best regards,
Stepan Seycek