Performance Monitoring and Management of Cloud Services

Performance monitoring and management of cloud services offer the basis to evaluate the quality-of-service and adoption benefits of cloud services and to correctly assess the current resource demands of an application.

Simply put, performance monitoring delivers the basic knowledge required to make scaling decisions confidently. In this section, we will discuss five cloud monitoring and management solutions such as :

  • Amazon CloudWatch
  • Private Cloud Monitoring Systems (PCMONS)
  • Cloud Management System (CMS)
  • Runtime Model for Cloud Monitoring (RMCM)
  • Flexible Automated Cloud Monitoring Slices (Flex-ACMS).

Amazon CloudWatch

Amazon CloudWatch is a proprietary monitoring solution for Amazon Web Services (AWS). This solution makes easy management of basic metrics such as storage and processing possible.

Further, it offers various types of statistics and self-configuration. It is designed specifically for managers and users of Amazon clouds and AWS products only.

Examples of similar commercial proprietary cloud monitoring solutions include Rackspace cloud monitoring, AccelOps, Zenoss, CopperEgg, and Monitis.

Private Cloud Monitoring Systems (PCMONS)

PCMONS is an open source monitoring solution developed for private clouds. It uses an integration layer to grant homogenous access to managers (administrators, services providers, cloud service brokers, and so on) and uses (cloud service consumers) that manipulate resources in a cloud.

It is specifically developed for uniform monitoring of infrastructure, regardless of the type of hosted resource(s) in a cloud.

An advantage of using PCMONS is its ability to integrate with other monitoring solutions complementing its performance. However, it can only be manually configured to monitor private clouds that undermine its autonomy and scalability.

Cloud Management System (CMS)

CMS leverages on RESTful Web Services to provide monitoring services. The REST serves as a technology for designing monitoring elements. As a result, CMS is able to integrate with other technologies, services and solutions that use REST.

It also supports integration with other monitoring solutions via the REST interfaces.

Runtime Model for Cloud Monitoring (RMCM)

RMCM is designed to monitor resources through abstract models which allows possible homogenous handling of heterogeneous resources.

In this way, a unified mechanism to access different resources like virtual systems and platforms in the same manner is provided.

RMCM is a 3-tier composite cloud monitoring model described as model for operators, model for developers, and model for users due to its flexibility to generate customized models depending on the request of each agent that integrates a cloud.

However, it requires a constant update of monitoring resources in order to maintain the model consistency. Similar to PCMONS, RMCM also requires manual installation and configuration of specific agents which compromises cloud monitoring requirements such as migration, scalability and autonomy.

Flexible Automated Cloud Monitoring Slices (Flex-ACMS)

Flex-ACMS is a composite and comprehensive cloud monitoring solution resulting from a rich integrated set of monitoring solutions.

Unlike PCMONS and RMCM, Flex-ACMS can be automatically configured to provide dynamicity and flexibility to cloud providers, which in turn, offers improved operational efficiency of cloud providers such as in billing and SLA. However, Flex-ACMS is configured by cloud administrators based on rules defined to indicate the metrics that must be monitored on each cloud slice and check monitoring solutions that need to be used to monitor each slice.