The new Process Execution Engine offered by the gCube platform consumes, manages and expose computational resources offered by the native gCube platform, underlying Grid infrastructure as well as external Cloud infrastructures. The computation providers span a large technological range including Web Services of several paradigms (REST-WS, SOAP-WS, WSRF), technology specific binaries or script executables.
Facilitating the exploitation and hiding the complexities, the Process Execution Engine constitutes a mechanism capable of orchestrating flows of invocations on this large and diverse set of targets over a multitude of enabling infrastructures.
The overall objective is one of bridging the gCube with heterogeneous platforms such as the underlying EGEE gLite Grid , the Condor, the Hadoop and more. By employing an architecture that involves a number of abstractions (infrastructure adapters and storage providers) and a number of standards such as JSDL for the specifications of the operations, the Process Execution Engine can serve effectively its cause.
INSPIRE - PE2ng interoperation
One of the interoperability cases the gCube platform targets in the context of D4Science-II is that of interoperating and providing computational abstractions and resources to the Invenio software on top of which the INSPIRE community offers its services. This is also one of the most extensive and broad spanning cases the Process Execution capacities are exercised.