Resource Prediction#

Synonyms: Workload Prediction, Workload Forecast.

In Brief#

Resource prediction is the estimation of the resources a customer will require in the future to complete his tasks. This concept has a wide variety of application and it is particularly studied in the context of data centres management. When these forecasts are generated, historical and current data are utilised to predict how many resource units, which tools and operative systems and the number of requests are required to accomplish a task.

More in Detail#

A resource prediction ensures that the resource pool is proportionate to the workload. To accomplish efficient work scheduling and load balancing in cloud computing, accurate resource requests forecast is required. It is vital for a competitive service to ensure that resources are available to fulfil demand as it arises. Cloud computing companies want to preconfigure computers ahead of time in order to deliver a high Quality of Service (QoS), which includes low latency, high availability and high dependability. If demand can be precisely forecast, suppliers may expect greater resource usage and a reduction on pre configured but idle computers, in addition to a high Quality of Service.

Cloud service demand, on the other hand, is difficult to forecast due to factors such as variety, size, burst and uncertainty. Service providers would be able to make a principled trade-off between QoS and resource cost if they had an accurate model of demand variance across time. The topic is widely studied in the literature and includes many algorithms and techniques applied over more than a decade, from statistical models to Machine Learning and Deep Learning models.

The benefits of predicting the future demands include better resource utilisation and a reduction of the overall location with the opportunity of serving more customers, which leads to an increase in profit and an overall reduction of energy and maintenance costs, with a possible improvement in environmental impact.

This entry was written by Andrea Rossi, Andrea Visentin and Barry O’Sullivan.