Cloud Capacity Planning

For available resources, capacity planning seeks a heavy demand. It determines whether the systems are working properly, used to measure their performance, determine the usage of patterns and predict future demand of cloud-capacity. This also adds an expertise planning for improvement and optimizes performance. The goal of capacity planning is to maintain the workload without improving the efficiency. Tuning of performance and work optimization is not the major target of capacity planners.

It measures the maximum amount of task that it can perform. The capacity planning for cloud technology offers the systems with more enhanced capabilities including some new challenges over a purely physical system.

The goal of Capacity Planners

The goal of capacity planners is to identify significant & vital resources that have resource ceiling & add more resources to move the restricted access to higher levels of demand. Network capacity is one of the hardest factors to resolve & the performance of the network is affected by I/O of the network at the server & network traffic from cloud to ISPs (Internet Service Providers).

Capacity planners try to find the solution to meet future demands on a system by providing additional capacity to fulfill those demands. Capacity planning & system optimization are two both different concepts, and you mustn't mix them as one. Performance & capacity are two different attributes of a system. Cloud 'capacity' measures & concerns about how much workload a system can hold whereas 'performance' deals with the rate at which a task get performed.

Capacity Planning Steps

  • Determine the distinctiveness of the present system.
  • Determine the working load for different resources in the system such as CPU, RAM, network, etc.
  • Load the system until it gets overloaded; & state what's requiring to uphold acceptable performance.
  • Predict the future based on older statistical reports & other factors.
  • Deploy resources to meet the predictions & calculations.
  • Repeat step (i) through (v) as a loop.

Cloud Application Baseline

The first thing that strikes in mind while dealing with the business issue is the system's capacity or working load as a measurable quantity over time, since many developers build their cloud-based applications & websites based on the LAMP. The full-form is extracted below:

  • Linux - operating system
  • Apache - Apache Software Foundation's Web-server
  • MySQL - database server
  • PHP - Hypertext Preprocessor

The above four technologies are open-source although the distribution may vary from cloud to cloud. There are other slight variations of the LAMP that are available for development. These are:

  • OpAMP (OpenBSD Apache MySQL PHP)
  • SAMP (Solaris Apache MySQL PHP)
  • WAMP (Windows Apache MySQL PHP)

Baseline Measurement

There are two important work-load matrices in the LAMP system. These are:

  • Page view: is the number of hits on a website & is measured in hits per second
  • Transactions: is measured by transactions per second and is the number of queries the database server completes per second

Load Testing

Server administrator checks for servers under load for system metrics to give capacity planners enough information to do significant capacity planning. Capacity planners should know about the increase in load to the system. Load-testing needs to query the following questions:

  • What is the optimum load that the system can support?
  • What system blocks the current system & limits the system's performance?
  • Can the configuration be altered in the server to use capacity?
  • How will the server react concerning performance with other servers having different characteristics?