In my first two blog posts, I gave an overview of what Session-only mode is and how to design your XenApp 6.5 enterprise farm using this model. In case you missed these posts, please see:
Speeding up Farm Deployments with XenApp 6.5 – Part 1

Speeding up Farm Deployments with XenApp 6.5 – Part 2

Today, I’d like to wrap up this blog series by sharing some test results with you that demonstrate the performance improvements seen when using XenApp 6.5 Session-only servers.

Test Results
The following tests compare the XenApp 6.0 IMA start time performance with a XenApp 6.5 Session-only server. For the test environment, a simulated WAN link was placed between the IMA data store and the member server. The WAN was configured for a 10Mb link with 200ms round trip latency. The base farm consisted of 50-servers and 100-applications.

The definitions of each test are as follows:

  • Farm Join: The time it takes for the IMA service to start when the XenApp server joins the farm for the first time.
  • LHC Sync: The time is takes for the IMA service to start and re-synchronize the LHC from the data store.

Figure 1. IMA service start performance for a single XenApp server over a WAN link


Figure 2. Data transfer usage to/from the data store during the IMA start process



Figure 3. Data store SQL round-trips during the IMA start process


To summarize, the Session-only mode feature in XenApp 6.5 provides:

  • Quicker deployment and provisioning of new XenApp servers to the farm
  • Reduced bandwidth and resource utilization of your IMA data store
  • Simplified management of deploying member servers in remote zones or sites

To learn more about Session-only mode, please refer to the Citrix XenApp 6.5 eDocs

Thank you for reading!