Keywords

1 Introduction

Cloud computing [24] is a techniques which allows suitable, on demand network access to the pool of various computing resources such as network, servers, storage application and other services, that can be quickly given back to the end user and released with minimal management effort. The cloud computing services [30] can be classified as Software as a Service Platform as a Service and Infrastructure as a Service along with different deployment models [8]. Essential characteristics of Cloud Computing [14, 27] are On-Demand Self-Service, Broad Network Access, Resource pooling, Rapid elasticity, Measured service and metering and billing.

Table 1. The Literature survey on Monitoring in to the cloud computing for efficient resource utilization
Full size table
Fig. 1.
figure 1

Cloud monitoring necessity

Full size image
Table 2. The Literature survey on prediction in to the cloud computing for efficient resource utilisation
Full size table

The resource management has to be efficiently used in the IaaS level of cloud computing, because the resources has to be allocated in a right amount [36] for an application. The interconnected resource management areas for efficient resource management are [5], resource discovery, Resource modeling, resource prediction and resource monitoring [22]. The Literature survey on Monitoring in to the cloud computing for efficient resource utilization are as follow (Tables 1 and 2).

Fig. 1 shows [19] the relationship between the cloud properties to the below mentioned key points as, scalability depends upon the aggregation of measures and filtering of measure etc.

2 Prediction

There are various case studies [20, 25] that indicates that the workload prediction plays very important role for any company. Basic Steps Required for Workload Prediction [2] and understanding the environment where the workload has to be executed, characterize the workload based upon its availability of resources or capacity planning, behaviour pattern etc., are key features for the prediction mechanism. The next step is to identify the parameter of the workload modelling, which depends upon the type of applications [2]. Lets now analyze the literature review of various prediction techniques.

3 Combining Monitoring and Prediction Techniques [10]

The above Fig. 2 indicates the connectivity between the monitoring and prediction mechanism to the cloud scenarios, their relationship has been identified [37] Prioritization Engine. As we can have more task aligned in the message queue in the cloud computing for different clients asking for the same resources in such scenario Business policies defined by the MSP (Managed Service Provider) helps to identifying the requests whose execution should be prioritized with respects to the services that they want, in case of resource contentions [15]. The rules engine evaluates the data captured by the monitoring system [33]. Rules engine and the operational policy is the key to guaranteeing SLA agreements.

Fig. 2.
figure 2

The combined architecture

Full size image

Monitoring System. Monitoring system collects the defined metrics in SLA. These metrics are used for monitoring resource failures, evaluating operational policies as well as for auditing and billing work. Monitoring system have to interact [6] with the other systems to optimise the its objective if careful usages of the resources available at IaaS of cloud.

Auditing. The adherence to the predefined SLA needs to be monitored and recorded. It is essential to monitor the values of SLA, as any noncompliance leads to strict penalties.

Prediction System: From auditing, the prediction model can be derived, which can be used to predict the resource consumption into the cloud and prediction can be merged with the Machine learning capabilities to increase its effectiveness.

Accounting/Billing System: Based on the payment model and metering The outcome of this model will predict the accurate resources usage for the specific type of work load and the problems that arises because of the under provisioning and over provisioning can be avoided.

4 Conclusion

In this survey paper we have highlighted the concepts of monitoring and prediction, which are an essentials for cloud computing environment, as the IaaS gives us a vision of infinite pool resources and managing such a huge resources while serving at local level as well as at remote level site is a tedious task and can be handled efficiently if the mechanism of monitoring and prediction concepts should be mapped to the cloud environment. The algorithms related to these two can be merged with the techniques of mathematical modelling, artificial intelligence and machine learning for better accuracy of results and analysis. The model which has been discussed at the concluding portion in this paper will allow the researchers to impose the techniques to implement monitoring and prediction at the correct position with respect to its associated attributes of the cloud computing environment.