If your scheduling and email services run on the Windows Server platform, chances are you’ll be using Microsoft Exchange. As a systems administrator, it’s your responsibility to make sure that things run smoothly. If you know what to watch out for, your job will be that much easier. This guide should help.
What is Microsoft Exchange?
Microsoft Exchange Server is a network server providing email and calendar functions for the Microsoft Windows Server product range. Exchange uses a collaborative approach to communications.
Exchange Online is a part of Microsoft Office 365, and some email hosts allocate Exchange accounts for home or personal use.
Design Your Message Infrastructure
Though designed to simplify matters, Exchange Server 2013 may itself consume considerable amounts of memory, CPU and disk storage capacity. So you’ll need to factor this in to your overall system design.
Performance bottlenecks, unscheduled downtime and erratic behaviour from system components may be minimised or avoided if you invest some time to plan, size and properly configure your messaging infrastructure.
Exchange 2013 Server Role Requirements Calculator is a tool that will help you determine the correct size for your Exchange Server hardware components.
To test your layout beforehand, the Microsoft Exchange Server Jetstress 2013 Tool simulates disk I/O load fluctuations on your server to ensure the stability and performance of your disk subsystem.
Set Performance Thresholds
From the outset, you should establish some long-term performance thresholds for your servers. These might include processor use, disk I/O, and RPC latency. Threshold values 
should be set to correspond to abnormal operating conditions (“red flags”). These thresholds will be specific to your organisation and its unique set-up.
Configure Your Storage
Examine the logical counters on each individual disk, to monitor latency. This should take in the total number of disk transfers each second that occur on your data devices, and the I/O activity levels for any drive supporting a database.
By making sure that requests occur within a set time period, you can establish how efficiently the read and write requests associated with a given disk are being handled.
It’s The User Experience That Counts Most
Focusing on system components is all well and good, but it’s your users who suffer if a failed service means they can’t access their mail. It’s the impact on your end users that should determine the focus of your monitoring activities.
A Fresh Outlook
Tweaking your server performance may have a knock-on effect. At the client end, having to deal with increased volumes of offline messages in a cached mailbox may cause problems for Microsoft Outlook – especially older versions of the program.
Particularly with Exchange 2013, you should check that the various client-side versions of Outlook seeking access to your servers are fully supported. As a blanket measure, you’ll typically need to exclude older versions entirely.
Managed Availability in Exchange 2013
Managed Availability is an on-board monitoring tool of Exchange 2013 which performs a set of periodic probes to monitor for potential sources of failure. The tool includes remedial measures which can be taken automatically in such cases.
Managed Availability is active monitoring. If a service fails for example, the tool will restart the service – which, to a passive monitoring tool might appear as a total collapse, until the rebooted service is running again. The remedial actions of Managed Availability mean that a mixed approach to monitoring is often best.
Active & Passive: Combined Monitoring
By combining observations from passive and active monitoring tools, you’ll get a combined view that’s better than both. Long-term collection of monitoring statistics can yield information that points to underlying issues. And by studying this data you’re in a position to make informed judgements about what’s really going on.
Monitoring Hybrid Deployments
Hybrid systems may include Exchange-specific elements related to federation and Mail flow, but other deployments might involve Directory Synchronisation (DirSync) and / or Password Synchronisation (Password Sync).
If your authentication protocols use Active Directory Federation Services, you’ll need to take this into consideration when monitoring your end users, as well. And with each Cumulative Update, new software features and functions become available.
All of this can be a real headache for traditional monitoring set-ups, as additional layers of complexity are added – to say nothing of more and more components you’ll have to monitor.
Your Tool-kit
As you’d expect, Microsoft ships a number of dedicated monitoring tools for its own brand Exchange Server.
The Office 365 Best Practices Analyser for Exchange Server 2013 has been part of the Exchange Admin Centre since the release of Service Pack 1. The Analyser scans Exchange servers to identify issues that don’t conform to Microsoft’s recommended practices. Scan results may be exported as HTML files.
The Exchange Server Deployment Assistant supports On-Premises, Hybrid, and Cloud Only deployment scenarios. It’s a form-based tool where you can enter basic information about your existing or planned system, and receive technical advice and feedback.
The Exchange Client Network Bandwidth Calculator can help predict your network bandwidth requirements for a specific set of clients.
At the third party level, ENow’s Mailscape for Exchange Online offers a complete solution for monitoring hybrid Exchange deployments. The market in this area continues to grow, so other products may soon become available.
Kerry Butters
Kerry is a published author and writer on all things tech, corporate tech, data centres, SEO, webdesign & more for some of the world’s leading sites.