Worst Practice #10: Ignore management tools |
 |
By David Sengupta
07 Jun 2005 | SearchExchange.com |
|
|
Achieving a high availability and well-performing messaging infrastructure -- whatever that means for your company -- requires that you understand your Exchange infrastructure. Monitoring solutions such as Microsoft Operations Manager 2005, combined with the Exchange Management Pack will help you ensure health by focusing on critical system events (i.e., from event logs and perfmon counters).
Reporting and analysis solutions, such as the Exchange Reporting Management Pack, provide the in-depth traffic analysis you need to ensure your Exchange transport infrastructure is performing optimally. Diagnostics solutions, such as the Spotlight on Exchange Management Pack, provide rapid troubleshooting and drilldown to the root cause behind any issues detected by MOM.
Archival solutions control the amount of data in your production Exchange environment and provide searchability of Exchange data, in case you need to respond to compliance-related or other investigations -- or simply find that lost e-mail.
Finally, you need to remember that Exchange performance also depends heavily on your network infrastructure (switches, hubs, VPN servers, firewalls, proxies, ISA servers, etc.), directory subsystem (domain controllers, Active Directory, global catalog servers, DNS, etc.), IIS subsystem (WWW service, SMTP service, IIS metabase, etc.), and e-mail client type (i.e., Outlook 2003 cached mode, Outlook Express, Entourage, etc.).
Operating Exchange without management tools is akin to driving a car without a speedometer or gas gauge. If you want to ensure the car gets you where you need to go, you have to keep an eye on those gauges.
If you want to ensure optimal Exchange messaging system performance -- and understand significant aspects of your Exchange infrastructure so that you can respond to those calls claiming that "e-mail is down" -- you need to be proactive about managing Exchange. Period.
Top 10 Exchange performance worst practices
 Home: Introduction
#1: Treat "high availability" as a future project
#2: Leave "IOPS" for the consultant
#3: Use identical configuration for all Exchange Server roles
#4: Encourage users to keep everything in their Inboxes
#5: Schedule backups and system maintenance during peak usage
#6: Throttle the RAM available to Exchange
#7: Virus scan and back up the M drive
#8: Ignore client configuration, type and usage
#9: Don't use change control
#10: Ignore management tools
| ABOUT THE AUTHOR: |
|
David Sengupta, Exchange expert
David Sengupta is a Product Manager in the Windows Management group at Quest Software. He has also been a Microsoft MVP in the Exchange Server category for six consecutive years. Sengupta has contributed to various Exchange and Windows books, magazines and white papers from a number of publishers. He also frequently represents Microsoft on staff at Ask the Experts, Microsoft Experts Area and Peer Talk at conferences such as MEC and TechEd. David has an M.T.S. from Tyndale Seminary, Canada, a B.Sc. from University of Ottawa, Canada and MCSE (Messaging) and CCA certifications. David runs a blog on Microsoft Exchange and e-mail compliance issues at http://p0stmaster.blogspot.com and can be reached at mailman@quest.com.
|
|
|
|
|
|
