In my Private Cloud talks with IT Pros, how best to virtualize Microsoft Exchange workloads, regardless of underlying hypervisor, has been a recurring hot topic area of discussion.  At Microsoft Tech Ed 2012, Jeff Mealiffe, Senior Program Manager on the Exchange team and responsible for Exchange virtualization guidance, delivered a great session on "Best Practices for Virtualizing Microsoft Exchange Server 2010".  This is a great resource to study when planning virtualized Exchange deployments! Although the session was specific to Exchange 2010, much of the overall guidance can be applied to Exchange 2013 as well. Below, I've included the following items to help you get started with virtualizing Exchange:

  • The recorded session video
  • A link to the downloadable slide deck
  • An indexed recap of my session notes
  • Links to additional tools and resources that I've personally found helpful

I'm definitely looking forward to building my next virtualized deployment of Exchange 2010 on Windows Server 2012 RTM when it releases on Sept 4th!  The increased VM resource densities in Hyper-V v3 of up to 64 virtual processors and 1TB RAM per virtual machine will be a big boost to virtualizing mission critical heavy-duty workloads like Exchange mailbox server roles and multi-role servers.

HTH,

Keith 

Best Practices for Virtualizing Microsoft Exchange Server 2010

Download a copy of the session slide deck.

Session Notes with Video Index

  • Supported Exchange Virtualization Scenarios [ 4:00 ]
  • Exchange 2010 SP1 or later
  • Hyper-V or any hypervisor in the Server Virtualization Validation Program (SVVP) - link provided below.
  • Items Not Supported when Virtualizing Exchange [ 7:00 ]
  • Hypervisor snapshots
  • Differencing / Delta disks
  • CPU oversubscription in a ratio > 2:1
  • Applications running on the parent / root partition
  • VSS backups of VMs from root
  • NAS storage of virtual disk files
  • JetStress Support in Virtualized Environments [ 12:10 ]
  • Supported in VMs on Microsoft Windows Server 2008 R2 or later
  • Supported in VMs on Microsoft Hyper-V Server 2008 R2 or later
  • Supported in VMs on VMware ESX 4.1 or later
  • More Info - http://bit.ly/HP8G0f
  • Big Problems to Avoid for Production Exchange VMs [ 17:53 ]
  • Dynamic Memory / Memory Overcommit  [ 18:00 ]
  • VM Snapshots [ 31:57 ]
  • CPU Oversubscription [ 35:05 ]
  • Overview of Best Practices [ 38:05 ]
  • Hypervisor adds CPU overhead - 10-12% in our Exchange 2010 tests [ 39:22 ]
  • Size for physical and provide those resources to each VM [ 40:31 ]
  • Exchange is architected for scale-out scenarios, avoid "all eggs in one basket" [ 40:48 ]
  • Resource Sizing [ 43:16 ]
  • Start with physical sizing process - use calculator (listed below)
  • Account for virtualization overhead (10-12%)
  • Determine VM placement to account for HA
  • Size root servers, storage and network infrastructure
  • Guest VM sizing [ 47:18 ]
  • Size Mailbox role first - other role sizes factored from Mailbox server requirements
  • Considerations for use of Multi-role servers - Mailbox, Hub and CAS roles on single VM
  • Unified Messaging Sizing [ 49:08 ]
  • Min 4 Virtual Processors (VP)
  • VM with 4VP & 16GB memory can handle 40 concurrent calls with Voice Mail Preview (65 calls without)
  • Storage Decisions [ 52:47 ]
  • Exchange storage separate from Guest OS virtual disk physical storage
  • Must be fixed virtual disk, SCSI pass-through (RDM) or iSCSI (terminated at host or guest)
  • SCSI pass-through (RDM) recommended to host queues, DBs and logfile streams unless using Hyper-V Live Migration where CSV is recommended
  • Must be block-level storage - NAS volumes not supported
  • Virtual Processors [ 56:04 ]
  • Prefer smaller number of multi-core VMs vs many single-core VMs
  • Don't assume that a hyperthreaded (SMT) CPU is a full CPU core
  • Private Cloud [ 57:08 ]
  • Good model for providing virtual infrastructure resources to Exchange, but be careful with "dynamic" cloud capabilities
  • Be prepared to apply different resource management policies to Exchange VMs
  • Host-based Failover Clustering [ 59:41 ]
  • Not an "Exchange Aware" HA Solution - Does not provide HA in the event of storage failure / data corruption
  • If using, combine with DAG when possible to provide maximum HA - Admin can re-balance DAG after failover to redistribute
  • VM Live Migration and Exchange [ 01:04:50 ]
  • DAG does not need to be dynamically re-balanced
  • Use CSV rather than pass-through LUNS for all Mailbox VM storage
  • Consider relaxing cluster heartbeat timeouts (5 seconds = default, 30 seconds = max recommended)
  • Size network appropriately for Live Migration
  • VM Placement [ 01:08:26 ]
  • Don't co-locate DAG database copies on same physical hosts
  • Distribute VMs running same roles to different physical hosts
  • If not using multi-role VM's, consider isolating mailbox and hub role VMs on separate physical hosts if possible

Additional Tools and Resources

  • HP Configurations
  • DELL Configurations
  • Unisys Configurations
  • EMC Configurations