UPDATE: Full proceedings of the 2015 workshop edition are now available online
Live blogging from the CloudMDE workshop, a workshop to explore combining model-driven engineering and cloud computing.
CloudMDE is an international workshop that aims to bring together researchers and practitioners working in MDE or cloud computing, who are interested in identifying, developing or building on existing synergies.
We aim to identify opportunities for using MDE to support the development of cloud-based applications (MDE for the cloud), as well as opportunities for using cloud infrastructure to enable MDE in new and novel ways (MDE in the cloud).
We also are interested in novel results of adoption of MDE in cloud-related domains, that provide insight into early adoption of MDE for building cloud-based applications, or in terms of deploying MDE tools and infrastructure on “the cloud” (MDE for the cloud).
Check the full program .
Let’s see what are the new research results in this area (you can also read the summary of last year’s edition, where we ended up proposing not less than 6 different DSLs for cloud modeling!):
Robert Crocombe and Dimitris Kolovos: Code Generation as a Service
Lowering the entry barrier for developers, specially by simplifying the distribution of code generators and ensure all members of the team are always working with the latest version of the generator. ANT is used to describe the generation task we would like to execute. Execution is asynhcronous to avoid server overloading. Generators are linked tom github projects which heavily simplifies the release and deployment of new versions of the generator.
Carlos Carrascal Manzanares, Jesus Sanchez Cuadrado, Juan De Lara: Building MDE cloud services with Distil
Motivation: How to facilitate the construction of MDE services, able to run in a cloud architecture?. To help on this they built Distil, a DSL to describe MDE services. With it you can describe the structure of MDE artefacts and the services. With that, Distil generates persistence services to store those artefacts, REST-based services to manage them and the project ready to be deployed in Heroku. Challenge: how to come up with dynamic realocation strategies for changing workload conditions.
Ta’id Holmes: Facilitating Migration of Cloud Infrastructure Services
Goal: evolve current forward-engineering approaches for provisioning of infrastructure services including an incremental evolution. This implies going for a model-based roundtrip engineering approach that operates on a diff-model. Appplicable to multi-cloud environment. These differences are then translated into a set of added/removed API calls wrt the original model generation.
Jozsef Makai, Gabor Szarnyas, Istvan Rath, Akos Horvath and Daniel Varro: Optimization of Incremental Queries in the Cloud
Using a RETE network to propagate changes that will then enable incremental queries. This is executed over a cloud infrastructure to achieve scalability, sharding separately dataa, indexers and the query network. This requires an architecture model specified with a textual DSL to specify virtual machines and node allocation. Allocation can be optimized for query performance. And configuration scripts can be code-generated from the DSL.
Francesco Basciani, Juri Di Rocco, Davide di Ruscio, Ludovico Iovino and Alfonso Pierantonio: Model repositories: will they become reality?
There’s an increasing demand for reusable modeling artifacts for benchmarking and learning purposes. Current support is very limited. This contrast with other disciplines (e.g. biomodels with 200k models, CellML, DDMore..). There are some technical challenges (variety of artefacts, relations among them, searching,…) but also non-techincal ones (reward mechanisms, licensing, moderation,…). They try to address some of these challenges in their MDEForge.
Alexander Bergmayr, Alessandro Rossini, Nicolas Ferry, Geir Horn, Leire Orue-Echevarria, Arnor Solberg and Manuel Wimmer: The Evolution of CloudML and its Applications
CloudML was the first (or one of the first) cloud modeling languages. A number of projects have worked on its (divergent) evolution. For instance, we have UML-based + profiles cloudML, external DSLs like MODACloudML, integration of different DSLs into a family of languages called CAMEL,… To make things even worse, no common operational semantics of CloudML are available.
Christoforos Zolotas and Andreas L Symeonidis: Towards an MDA Mechanism for RESTful Services Development
Proposal of a new language to model RESTFul services going beyond typical non-CRUD operations, which is the only part that other languages basically cover. The current version of the language and tool support is available on GitHub.