The first and foremost is availability of matching resources. Another important parameter is proximity to the data, if distributed data processing is involved. However, since it is difficult to have up-to-date information and uniform priorities in Grid, the latest approach is to schedule generic tasks that pull payloads from a central queue after analyzing the environment in which they ended up.
the Grid environment is changing dynamically in time mainly due to load of the system, available network bandwidth and also due to other events such as resource or network failure or recovery.
thank u sir...but sir can u please elaborate me job failure rate in which sense. Plz also forward me the link of the research paper carried on the same topics in grid environment
On of the challenges of grid computing is to execute jobs on heterogeneous machines without much control over the machines availability. (machines can be removed from the grid at any point of time without warning).
Thus, the key parameter would be ensuring good performance on heterogeneous environment and jobs continuity in the event of machines failures.
you can use parameters like throughput , Reliability ,Durability ,Agility , Security . you can add parameters according to your needs and challenges for your Research Work .
Grid environment is heterogeneous in nature and it is dynamic (resource can join or leave at any time). In scheduling you have to consider this and you can use parameters like makespan, reliability, failures ....