Eyal Pozner, M.Sc. Thesis Seminar
Wednesday, 10.4.2013, 13:30
Physical memory is the most expensive resource in use in today's
cloud computing platforms. Cloud providers would like to maximize
their clients' satisfaction by renting precious physical memory to
those clients who value it the most. But real-world cloud clients
are selfish: they will only tell their providers the truth about how
much they value memory when it is in their own best interest to do
so. Under these conditions, how can providers find an efficient
memory allocation that maximizes client satisfaction?
This research presents Ginseng, the first market-driven framework for efficient
allocation of physical memory to selfish cloud clients. Ginseng
incentivizes selfish clients to bid their true value for the memory
they need when they need it. Ginseng continuously collects client
bids, finds an efficient memory allocation, and re-allocates physical
memory to the clients that value it the most.
During the research, a new type of application was evolved, to be used
efficiently in a system with dynamic memory conditions. A special benchmark was
developed among a modification of a widely used caching application called
memcached. It was also necessary to find special configuration of the OS
allowing high memory percentage usage without the OS interference.
A whole environment was developed around Ginseng, for experimenting and testing
the it with different workloads, for simulating it under different conditions,
and for comparing the results to determine the system efficiency.
Ginseng achieves a x6.2-x15.8 improvement in aggregate client satisfaction
when compared with state-of-the-art approaches for cloud memory
allocation. It achieves 83%-100% of the optimal
aggregate client satisfaction.