Browse Prior Art Database

Network Attached Storage Capacity Planning Tool

IP.com Disclosure Number: IPCOM000015733D
Original Publication Date: 2002-May-08
Included in the Prior Art Database: 2003-Jun-21
Document File: 7 page(s) / 126K

Publishing Venue

IBM

Abstract

Network Attached Storage Configuration Tool Algorithm

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 22% of the total text.

Page 1 of 7

Network Attached Storage Capacity Planning Tool

Network Attached Storage Configuration Tool Algorithm

Disclosed is an algorithm that can be used to determine a reasonable configuration that provides optimal performance for Network Attached Storage (NAS) solutions. The tool is based upon selecting a NAS configuration based upon measured results obtained from industry standard benchmarks such as Ziff-Davis NetBench . Therefore, this algorithm is based upon an assumption that the NAS workload will be typical Windows office application file server usage. The algorithm currently does not consider server-attached storage. Server attached storage configuration is significantly more complex and beyond the scope of this disclosure.

The first step of this procedure is to interview the customer to determine the type of network connection and number of users the customer would like to support with the NAS solution. The second step is to use the NAS benchmark results that have been determine by laboratory measurements to extrapolate a suitable NAS configuration for the desired number of users. The extrapolation is based upon several important characteristics of NAS operation. These are:

1). A NAS cannot supply more data throughput than that provided by the total LAN connection to the NAS.

2) Each active user requires a predetermined amount of memory for buffer space on the NAS for file caching.

In addition, several input parameters are required to generate an optimal NAS configuration. These are:

1) Maximum sustained NAS I/O bandwidth expected per user (a default is listed in case the user does not know this.) However, it is a simple matter for the user to measure maximum sustained throughput for a few existing power users, and to use this value as a basis for future configurations.

2) Maximum number of users that will store data on a single disk (a default is listed in case the users does not have this information.) This value is used to help prevent disk bottlenecks that would occur due to excessive seek commands generated by too many users accessing data stored on too few disk drives. Again, a default value is provided but an override mechanism is provided should the user have experience to suggest a more suitable number of users per disk than the suggested default. This value can be approximated by examining an existing fileserver or NAS solution which provides adequate performance. Maximum users per disk can be determined by dividing the total number of active by the total number of disks.

3) Performance for each NAS model to be recommended must have been measured using a suitable file server benchmark such as NetBench. The benchmark results are used to determine

1

Page 2 of 7

the maximum sustain data throughput of each model for typical office file server workloads.

Finally, the tool provides a mechanism for the customer to enter an expected growth rate for the user community over the life of the NAS solution. This enables the tool to...