...

• What bottlenecks exist in the functionality of (for example) storage, access and delivery of data, data processing, and workflow management?
• What are the current peak volumes for data access, storage and delivery for parts of the infrastructure?
• What is the (computational) complexity of different data processing workflows?
• What are the specific quality (of service, of experience) requirements for data handling, especially for real time data handling?

Optimisation requirements gathering is coordinated by Image Removed with help from go betweens.

Overview and summary of optimisation requirements

...

Streamlining the acquisition of data from data providers is important to many RIs, both to maximise the range and timeliness of datasets then made available to researchers, and to increase data security (by ensuring that it is properly curated with minimal delay).

Optimisation methodology

Knowledge infrastructure

The efficient exploitation of RI infrastructure, in terms of data transportation, placement, and the serving of computing resources, requires knowledge about the RI and its assets. This knowledge is usually embedded in the technical experts assigned to manage the infrastructure; however the ability for infrastructure services to acquire the knowledge to manage themselves (even if only to the extent of provisioning resources on cloud infrastructure to support static resources) would allow for greater flexibility and agility in RI compositionOptimisation of infrastructure is dependent on insight into the requirements and objectives of the system of research interactions that the infrastructure exists to support. This insight is provided by human experts, but in a variety of different contexts:

• Concerning the immediate context, the investigator engaging in an interaction can directly configure the system based on their own experience and knowledge of the infrastructure.
• Concerning the design context, the creator of a service or process can embed their own understanding in how the infrastructure operates.
• Alternatively, experts can encode their expertise as knowledge stored within the system, which can then be accessed and applied by autonomous systems embedded within the infrastructure.

In the first case, it is certainly possible and appropriate to provide a certain degree of configurability with data processing services, albeit with the caveat that casual users should not be confronted with too much fine detail. In the second case, engineers and designers should absolutely apply their knowledge of the system to create effective solutions, but should also consider the general applicability of their modifications and the resources needed to realise optimal performance in specific circumstances. It is the third case however that is of most interest in the context of interoperable architectures for environmental infrastructure solutions. The ability to assert domain-specific information explicitly in generic architecture and thus allowing the system to reconfigure itself based on current circusmtances is potentially very powerful.

If one of the goals of ENVRI+ is to provide an abstraction layer over a number of individual research infrastructures and a number of shared services that interact with the majority of those infrastructures to provide standardised solutions to common problems, then the embedding of knowledge at every level of the physical, virtual and social architecture that is a necessary result of this approach is essential to mediate and optimise the complex system of research interactions that are then possible.

Research Infrastructures

...