 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
Windows Server Forum / Exchange Server / Design / July 2006Performance Exchange 2003 Standard vs Enterprise
Performance Exchange 2003 Standard vs Enterprise Hi, We are now running an Exchange 2k3 Enterprise cluster but I need to add more servers as performance drops with 700+ users and ever growing mailboxes and e-mail usage. To give you an idea of the size: our total Exchange store size is about 70 GB. Maibox sizes are generally around 150MB with a few up to 2 GB (yes. I know.) and may users also use PST files located on the File server. We already have an network load balanced FrontEnd (Standard Edition). I have to add one or more new servers to improve performance and want to migrate the existing servers to a VMWare ESX environment (hooked up to a high end SAN) as well. The Exchange Virtual servers would be running on a IBM X366 3.16 GHZ Xeon server and be given each 2 processors and as much memory as needed. What would give a better performance: - 2 Enterpise Edition servers (not clustered) using all storage groups to keep the stores relatively small - 3 or 4 Standard edition servers each hosting 1 store Up to SP2 we would have to opt for the Enterprise solution, but now the maximum store size is extended in the Standard edition it's a different story. Obviously there is also a licensing cost aspect to the equation, but at this point I'm wondering what would be the technical best solution. I spend quite some time searching the web for performance differences between the Standard and Enterprise Edition versions, but didn't find anything helping me make this decision. Many thanks, Michel Instead on cocentating on the version, I'd concentrate on resource bottlenecks. > Performance Exchange 2003 Standard vs Enterprise > [quoted text clipped - 38 lines] > > Michel Hi John, Connectivity or things like availablity of Global Catalog servers are good, that is not the problem. And with VMWare ESX it's easy to scale up the recources like processor usage or memory. I just would like to know if more small standard servers are preferable over fewer heavier Enterprise servers. Even the Microsoft System Center Capacity Planner 2006 doesn't say anything about Standerd or Enterprise edition.. Nor does it allow you to take this into account when you run the simulation. > Instead on cocentating on the version, I'd concentrate on resource > bottlenecks. [quoted text clipped - 41 lines] >> >> Michel Best technical solution? That would be to not use VMWARE to host the servers for starters. The two versions are binary compatible versions. One has more features than the other. If you would rather run several versions of standard on separate servers, then you'll get similar performance than if you ran multiple storage groups/db's on a single except that your network, cpu, storage and memory wouldn't be hit as hard on the standard boxes. 700+ mailboxes is really not that many in this day and age. Sub-1000 machines are not as common as they used to be in favor of 1500 and greater mailbox servers. The biggest factor affecting your db size (and hence your mailbox sizes etc) is the amount of restore time it requires. Larger db's take longer to get on and off tape than do smaller one's in multiple storage groups (which can be backed up and restored in parallel). There are plenty of sizing guides. Microsoft has a pretty good one at http://www.microsoft.com/exchange/library if you're interested. You can search the net for John's name and see plenty of sizing related to Exchange especially around disk sizing. He's very good at it and has posted consistently good information about it. VMWARE won't be able to scale as well with Exchange and your SAN may or may not do what you think it will. Be sure to check and to remember that Exchange is generally cache-unfriendly when it comes to SAN technology. There are better SAN algorithms now, but you'll want to be sure you're running SAN technology that takes that into account. > Hi John, > [quoted text clipped - 53 lines] >>> >>> Michel "and your SAN may or may not do what you think it will. " In my previous career, prior to IT, the mantra was "In god we trust, all others we verify" I guess this mantra carries over to IT: measurement of SAN IO response times would be a good idea. I've lost count of the number of times undersized SAN storage did not perform as expected. Again, I would concentrate on resource bottlenecks. John > Best technical solution? That would be to not use VMWARE to host the > servers for starters. [quoted text clipped - 83 lines] >>>> >>>> Michel |
Reply to this Message |
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2009 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.