Microsoft 1 manual Capacity and Availability Management

2

Capacity and Availability

Management

Introduction

In the vast majority of cases, the load on your Exchange 2000 Server computers will increase over time. Companies increase in size, and as they do, the number of Exchange users increases. Existing users tend to use the messaging environment more over time, not only for traditional e-mail, but also for other collaborative purposes (for example, voicemail, fax, instant messaging, video conferencing). The load on the messaging environment will also vary over the course of the day (for example, there may be a morning peak) and could vary seasonally in response to increased business activity.

The aim of your operations team should be to minimize the effect of the increased load on your users, at all times keeping within the requirements set by your service level agreement (SLA). You will need to ensure that existing servers running Exchange are able to cope with the load placed upon them (and upgrade hardware if appropriate).

Another important requirement of the operations team is to minimize system downtime at all times. The level of downtime your organization is prepared to tolerate needs to be clearly set out in the SLA, separated into scheduled and unscheduled downtime. Many organizations can cope perfectly well with scheduled downtime, but unscheduled down- time almost always needs to be kept to a minimum.

Exchange 2000 is predominantly self-tuning, but there are areas where tuning your servers running Exchange will result in an improvement in performance. It is important to identify these areas and tune where appropriate.

Inevitably there will come a point where the load on your servers running Exchange is such that hardware upgrades are required. If you manage this process effectively, you can significantly reduce the costs associated with upgrading.