Image Processing Reference
In-Depth Information
[,]
t d
SP
[,]
t d
SP
d
i
d
j
MN
m
Fig. 7.1 Cloud rack configuration: Mobile and static nodes are allocated on cloud rack and
resources are inter-exchanged among mobile and static users based on the best effort allocation
Fig. 7.2 Multi-migrate-resume scheduling cloud policy based on “a-priori” task partitioning
resource migration and offloading approach that aims to release an aggravated
terminal in data center rack, and migrate according to certain parameterized
requirements the resources to other racks based on the minimized service and
processing delay, respectively. This work takes into account the impact of different
files capacities, by using multi-migrate-resume scheduling policies under various
conditions. The aim of this study is to achieve high performance, while preserving
fairness in terms of sequential task execution-access (claims). Tasks consist of
sequential files that are uploaded and run on terminals on the cloud rack (Fig.
7.1
).
Tasks start to execute, if there are terminals on cloud rack that are available to
handle them. When a claim is large enough and the terminals of the cloud rack can
no longer handle the request of a mobile user, the multi-migrate-resume policy
(Fig.
7.2
) is used. This policy enables the claims to be partitioned “a-priori” to
