Released: v12.8 of the Exchange 2010 Mailbox Server Role Requirements Calculator


Important – We’ve released version 12.8 to replace versions 12.3 and 12.1 of the calculators due to bugs found. Please download the latest version.

That’s right, there are some serious updates to the Exchange 2010 Mailbox Server Role Requirements Calculator, including one piece of functionality that you all have been asking for since I released the first version back in November!

Please go to our Mailbox Server Role Storage Requirements Calculator updates tracking page to see what is in this new version!

A blog post explaining the calculator (updated for this new version) is here and or you can download the calculator directly.

Comments welcome!

Ross Smith IV

Comments (20)
  1. Robert Greenlee says:

    Unless I’m missing I don’t see any functionality to calculate tiering the archive databases off to seperate storage.  Is it on the plan to add that at some point?

  2. Exchange says:

    We’ve updated the calculator to v12.3 to fix an issue with respect to active/active (single DAG) and having a lagged copy for both the primary and secondary datacenters.

    Ross

  3. Exchange says:

    Robert, at this time there are no plans to add archiving tiering.

    Ross

  4. Robert Greenlee says:

    I guess I just have to run it twice.  Once to get DB design for the main databases and another for Archive.  Although it’ll probably be too late for us I hope you do add in to made the new functionality so many customers were begging for.  Thanks.

  5. Timothy McCarty says:

    Reviewing a current pass through the storage calculator it seems to me that there must be something wrong with the "Mailbox Role CPU Utilization" calculation.Using all of the same inputs I see that the "Primary Datacenter Server" utilization value drops from 74% in version 7.8 to 4% in version 12.3.  I would be happy to mail an example along if that would help.

  6. Jeff Guillet [MVP] says:

    Shweet!  Thanks, Ross.  Great job as usual!

  7. Timothy McCarty says:

    In looking for why this would be I see now that you changed what data is input for the CPU calculations; for others – where before you entered "Adjusted Megacycles /  Core" (ver 7.8) you now use the raw "SPECint2006 Rate Value" and the calculator does the math for you.  

    Sorry Ross!

  8. Dan Sheehan says:

    Ross… you are the man.

    Thanks for the update!

  9. littlecayman says:

    Great!

  10. Great!

    Note: you forgot to remove that 2nd DAG FSW in the 1st diagram.

  11. MattWade says:

    How would one go about determining the SPECint rate for virtualized mailbox servers?  The calculator makes no distinction between a physical or virtual server.  After much discussion among my peers this is what we came up with solution.  Are we on the right track?  

    Example: Our chosen host platform is an HP ProLiant DL585 G7 (2.3Ghz Opteron 6176) 48 Core.  Result value of 782.  

    The first step is to break down the server result value into an individual core result value.  We did this by taking the server result value and dividing it by the number of cores in the server.  Using our example server above the math would look like this:

    Server result value / # of cores = core result value

    782/48 = 16.29

    We would then take this new core result value and multiply it by the number of cores we are assigning each virtualized mailbox server.  In our case it is four.

    Core result value * # of cores per vm = adjusted SPECint Rate Value

    16.29 * 4 = 65.17

    We then take this new value of 65.17 and enter this value into the SPECint Rate Value field and set the cores/server variable to 4.

  12. @MattWade:

    The CPU overhead associated with running a guest operating system in a virtual machine was found in testing to range from 9 percent through 12 percent ..  We recommend reducing the user capacity of Mailbox servers by 10 percent to account for hypervisor processor overhead.

    http://technet.microsoft.com/en-us/library/dd346699.aspx

    I’d take that 782 figure, deduct that percentage and then do your calculations.

  13. @Ross & Co. : A "Virtualized yes/no" checkbox in the calc sheet would be nice to deduct the estimated overhead percentage automatically.

  14. Jason K says:

    Does anyone know if RIM published their Exchange 2010 guidance?

  15. Jim says:

    Hi Ross

    I cant seem to get the active/active configuration to work properly. We have a design that has 2 servers each with active mailboxes in both the PDC and the SDC so 4 in total with a single DAG. I have tried different inputs but the outputs for the SDC don’t seem to be complete. Can you advise in our scenario what we should be entering for the Number of Mailbox Servers Hosting Active Mailboxes / DAG (Primary Datacenter). We have the site resilliense set to Yes and ActiveActive. I seemed to of tried everything but cannot understand the results.

    Thanks

    Jim

  16. Jim says:

    Hey Ross

    Ignore that I figured it out in the end. I was missing mailbox database copy configuration information!

    Awesome tool!

    Thanks

    Jim

  17. Blacktoe3 says:

    I’m having issues with the computations for the number of cores per role.  I’m currently getting #DIV/0! for my values.  I have used a value of 266 for my SpecIntRate for an xx5570 Dell R710.  The calculator calculates the secondary datacenter with no issues, but it is not calculating the primary datacenter.  Help please!  Thank you,

  18. Blacktoe3 says:

    Looks like the calculator is requiring more than 1 Active server in the primary data center.  So, there is no way to have perhaps an active/passive at the data center, then a passive in a secondary data center?  And would I just double the calculation for the amount of cores if I decide to proceed this way?  Thanks again,

  19. DavidRa says:

    @Blacktoe3: It’s my ill-informed opinion that if you want A+P in the primary data centre and P in the secondary, set up the Mailbox Database Copy Configuration like this:

    Total Number of HA Database Copy Instances (Includes Active Copy) within DAG 3

    Total Number of Lagged Database Copy Instances within DAG 0

    Number of HA Database Copy Instances Deployed in Secondary Datacenter 1

    Number of Lagged Database Copy Instances in Secondary Datacenter 0

    Also set Exchange Environment Configuration:

    Number of Mailbox Servers Hosting Active Mailboxes / DAG (Primary Datacenter) 2

    ———-

    Am I the only one struggling with the description and data for the SPECint2006 Rate Value, when virtualizing? The comment (cell B113) suggests that the rate should be entered as a total into the Megacycles per Core field (which is now renamed to the SPECint2006 Rate Value cell, I think).

    But … if your mailbox server has only 4 CPUs because it’s a Hyper-V VM, and the physical host has (say) 12 + hyperthreading; do you divide the SPECint2006 rate by 3, 6, 12, or 24!? An example is the X5650 processor in a Dell M610 blade, which has a score of 339.

    Seems to me:

    * Divide by 3 if you ignore HT cores and are entering the total rate across 4 cores;

    * Divide by 6 if you include HT cores and are entering the total rate across 4 cores;

    * Divide by 12 if you ignore HT cores and are entering the rate per core;

    * Divide by 24 if you ignore HT cores and are entering the rate per core.

    So … which one is right?

    Thanks in advance.

Comments are closed.