On Tuesday Alex and I attended a workshop called Multi-Cloud Scenarios for the Future Internet organised by James Ahtes at Cloud4SOA. Camp and Cloudsoft excluded, the focus of the presentations was on managing interoperability between different clouds.
Here's a summary of the projects we heard about:
An HTTP protocol for cloud. Supports OpenNebula, OpenStack and - as revealed through the session - some of the projects below. Made a play of transparency in spec and organisation.
A toolkit that has providers install something (it was unclear) at the datacentre that allows developers to use the Optimis IDE (Eclipse plugin, I think) to select the best environment for their needs, based on Trust, Risk, Eco and Cost (TREC) parameters.
Similar to Optimis, allowed SLA and QoS management too. Only supports OpenNebula at the moment.
Different to the other EU projects, BonFIRE is a cloud designed for researchers and experimentation. Has some specialist networking features.
Multi PAAS management and portability. Matchmakes between clouds, an API layer 'harmonises' different providers' APIs. Defines SLA terms and monitors breakages. They will be open sourcing their work on GitHub imminently.
Also of interest was a PAAS Match project, mentioned briefly, that should also be released soon. I believe it was along the lines of 'I have these requirements, which clouds suit?'
Early days for a project that seemed very ambitious:
The main goal of MODAClouds is to provide methods, a decision support system, an open source IDE and run-time environment for the high-level design, early prototyping, semi-automatic code generation, and automatic deployment of applications on multi-Clouds with guaranteed QoS.
Flexiant are involved with this.
The Autonomic Manager will automate the elastic scaling of the Data Platform and orchestrate the self-optimizing strategies that will dynamically reconfigure the data distribution and replication mechanisms to maximize efficiency in dynamic workload scenarios.
It was interesting to hear about these large EU projects. They seemed a bit too determined to solve everything. They certainly made Camp look quite modest in its aims. Of all of them I'll be interested to check on the Cloud4SOA GitHub release when it happens. They're paused/stuck on the 'I'd like to tidy this up' stage. The brief mention of their PAAS Match project piqued my interest too.