Cloud Computing

Competency Leader

John Murphy

Cloud computing is resulting in a scalable infrastructure available to industry as a commodity-based service. Cloud computing provides for software, platform, and infrastructure as a service. Cloud computing is becoming a critical service for industry today with a large proportion of industry today availing of cloud services. ECS will apply a range of research outputs to the area of cloud computing. In particular, tools will be developed that can be used as part of testing and analysis of such systems, both during the development process and for maintenance purposes.

We have produced a white paper on Cloud Computing that outlines our vision for Cloud Computing. We have also identified a number of high-level research challenges that are listed here (in no particular order):

What advances are required to enable carbon-neutral Cloud Computing?
What is required to make data/applications truly portable across Cloud providers?
What techniques are needed to streamline the migration of applications so that developers can easily build, deploy, run, and monitor their existing applications in the Cloud?
Can we design tools to enable applications to automatically enable surge computation into the public Cloud when load is high?
Can we generate fully customisable SLAs?
Can we secure Cloud Computing from end-to-end?
Can we produce standards for Cloud computing to make data seamlessly portable between Cloud providers?
What tools can we deploy to enable developers to debug Cloud application software - taking into account of the characteristics of Cloud Computing, i.e., parallel, distributed, and potentially massive data throughput?
How can we scale up (and down) the provision of Cloud resources to a client invisibly?
How can we provide real-time billing to Cloud clients?
How do we effectively protect the reputations of Cloud tenants when some of those tenants are bad?
How do we verify that data is not leaking in multi-tenant architectures?
How should Cloud computing be licensed in the future to maximise the benefit to all?
How can we enable clients of Cloud Computing to determine which data is safe to store in the Cloud?
How do we load-balance Cloud computing at a global scale - accounting for differing power costs, client usage profiles, and resource availability?
How do we partition a global Cloud computing system such that private data remains within appropriate jurisdictions - at a national and international level?
How can we improve the throughput of data into and out of the Cloud? - will the network always be a bottleneck?
How can we model SLAs so that clients can easily understand how to specify their requirements from a Cloud provider? Can we ever get beyond the need for SLAs?
Can we make Cloud computing self-managing? Do what extent can we take the human out of the loop?
How can we ensure that SLAs are always enforced?
For ultra-reliable Cloud computing applications how can we use multiple Cloud providers together?
What controls are necessary to enable a Cloud provider to protect user data from security breaches? Can security be verified?

If you have any thoughts, comments, or suggestions on these challenges, or would like to suggest additional challenges please get in touch with Lorcan Coyle (Lorcan dot Coyle at lero dot ie).