If there any techniques or approaches to applying fairness when we allocate the resource in a cloud.
Hi,
Yes, of course, its an interesting research topic in the cloud, and some papers have benn published : look at this one for example : https://www.cs.berkeley.edu/~alig/papers/drf.pdf
It depends what you mean by resource: is it a VM or a physical machine running a VM? Depending on the choice, provider (for PM allocation in the datacenter and VM-to-PM provisioning) or client (task-to-VM scheduling) objectives can be considered. The answer is not simple as it depends on your objective(s): energy, load balancing, resource utilization, execution time, throughput, latency. Fairness is a vaguely defined term: for instance it can refer to "all users have the same experience in terms of response times". You can take a look at the numerous resource scheduling and allocation papers dealing with clouds. They all address fairness in terms of the objective(s).
I agree with you Marc.
By the way, are you conducting research works on this subject ? It sounds interesting for me. Dont hesitate to contact me if you want to collaborate with my team.
There are many available articles on resource allocation in cloud computing environments.
However, I would consider checking work with elasticity in cloud computing as they often load balancing situations and better resources allocation, both considering infrastructure aspects (CPU, memory, network, etc.) or applications/systems aspects (throughput, response time, etc.), helps in achieving fairness.
- Badges
- Science method