Browse Prior Art Database

Workload Capability for Cloud

IP.com Disclosure Number: IPCOM000239160D
Publication Date: 2014-Oct-17
Document File: 2 page(s) / 28K

Publishing Venue

The IP.com Prior Art Database

Abstract

Today, it is very difficult to understand the full capability of a specified Cloud. Even more, it is difficult for an application to determine if the cloud can Scale-Up or Scale-Out and if its workload will be well managed then. The main idea of my solution is to define new measurements and metrics and provide a method that will use it to determine the scalability and potential of a Cloud. These metrics and method could help then to evaluate a Cloud.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 53% of the total text.

Page 01 of 2

Workload Capability for Cloud

Today, it is very difficult to understand the full capability of a specified Cloud. Even more, it is difficult for an application to determine if the cloud can Scale-Up or Scale-Out and if its workload will be well managed then.

    The main idea of my solution is to define new measurements and metrics and provide a method that will use it to determine the scalability and potential of a Cloud. These metrics and method could help then to evaluate a Cloud.

    For example, what is the difference of the Cloud built on commodity servers and one built on High-End servers.

Description


Two types of Cloud have to be defined: · SUC: Scale-Up Cloud, based on High-End systems (Like System z or High-End Power).

· SUC: Scale-out Cloud, based on distributed Low-End systems (like Blades servers).

    Therefore any Scale-Up or Scale-Out application will be sent on the appropriate one, managing its workload in the best way.

Workload Capability concept

    
The following method will help them to determine the maximum workload capacity of the Cloud.

    First we need to define, as below, a new metric called WC for Workload Capability

· C = Number of Cores.

· P= Power in MIPS (Million Instructions Per Second)

· R= RAM in GB · S= Storage in TB · N= Network in Gb
WC = C.R.S.N
The WC can easily be represented by a table.

    Server specification: 2 sockets (8 cores), 25 MIPS, 256GB of memory, 10TB of storage, 10Gb network, the table (WC Table) will be the following:

CPU PWR RAM Storage Network 16 25 256 10 10

    We can also add other attributes like, redundancy, security, availability, etc. Server Workload Capability
The maximum Workload Capability of a server is then define by its WC and represented by the four attributes: C, P, R, S, N.

Example Server 1:

    
Server specifications: 2 sockets (8 cores), 10 MIPS, 16GB RAM, 1...