A VMware Horizon solution on Dell EMC PowerEdge R7525 servers based on 2nd Gen AMD EPYC processors
Mon, 12 Dec 2022 13:38:14 -0000
|Read Time: 0 minutes
Many VDI deployments experience performance issues and poor user experience when trying to maintain a cost-effective consolidation ratio. A higher consolidation ratio of virtual machines to physical servers offers better economics and lower Total Cost of Ownership (TCO). The amount of TCO benefits might vary depending on the size of your VDI environment. It is a challenge for today’s organizations to deploy a cost-effective VDI environment while striking the right balance between performance and density.
The Dell Technologies Ready Solutions for VDI team provides a solution that resolves these challenges. It uses VMware Horizon based on Dell EMC PowerEdge R7525 servers equipped with new 2nd Gen AMD EPYC processors. The PowerEdge R7525 is a highly scalable, two-socket 2U rack server that delivers powerful performance and flexible configuration options. The servers are equipped with 2nd Gen AMD EPYC processors that can accommodate up to 64 cores per socket. A dual-socket R7525 server can have up to 128 cores, providing excellent user densities and a lower TCO for your VDI deployment. This solution offers you the flexibility to correctly size your VDI environment for performance and an exceptional end-user experience.
In this blog, we will discuss the key benefits of this solution and the results of performance testing carried out by the Dell Technologies Ready Solutions for VDI team.
Key benefits of the solution
- High performance and density: PowerEdge R7525 servers based on 2nd Gen AMD EPYC processors are designed for performance and with a high number of cores per CPU socket you can achieve higher user densities per server.
- Lower security risks with a diverse CPU architecture: The 2nd Gen AMD EPYC processors in this solution present an opportunity to diversify the CPU architecture within your data center. A data center with diverse CPU architecture poses a lower risk to your organization during security threats. Customers can move business-critical data to an appropriate and safe environment while a security event is resolved. With AMD Infinity Guard, which includes the AMD secure processor, Secure Memory Encryption (SME), and Secure Encrypted Virtualization (SEV) capabilities, you can minimize potential attack surfaces and deploy your workloads with confidence.
- Excellent graphics capability: The solution also offers excellent graphics performance with the capability of hosting up to 6 NVIDIA T4 cards (each with x16 PCIe lane access) on the PowerEdge R7525 server, providing up to 96 GB of graphics frame-buffer per server.
Solution performance testing
The Dell Technologies Ready Solutions for VDI team used the Login VSI benchmark tool for performance testing. We performed testing on a 3-node VMware vSAN cluster based on PowerEdge R7525 servers with a ‘Density Optimized’ configuration. VMware ESXi 6.7 update 3 was used as the hypervisor and the Horizon 7 virtual desktops were provisioned by instant clones. See Figure 1 for the solution stack.
Figure 1: VMware Horizon on PowerEdge R7525 solution stack
The environment configuration was:
- PowerEdge R7525 server (Density Optimized configuration)
- 2 x AMD EPYC 7502 (32 core @2.5 GHz)
- 1024 GB (16 x 64 GB @ 3200 MHz)
- 2 x 800 GB WI SAS SSD (cache)
- 4 x 1.92 TB MU SAS SSD (capacity)
- Mellanox Connect X- 5, 25 Gbe Dual port SFP28
- 6 x NVIDIA T4
- vSAN all-flash datastore
- VMware ESXi 6.7u3 hypervisor
- VMware Horizon 7.10 VDI software layer
See Table 1 for the VM configuration that we tested for different Login VSI workloads. For details of the test environment, configuration and testing process and an analysis of the test results, see the Reference Architecture Guide available on the Dell Technologies VDI Infohub.
Table 1 : Virtual machine configuration for different Login VSI workloads
Figure 2 shows the recommended density figures per host for Login VSI workloads based on our performance testing. We recommend these density figures after monitoring and analyzing a combination of host utilization parameters (CPU, memory, network and disk utilization) and Login VSI results. We monitored the relevant host utilization parameters and applied relatively conservative thresholds for the Login VSI testing. Thresholds are carefully selected to deliver an optimal combination of excellent end-user experience and cost-per user while also providing burst capacity for seasonal or intermittent spikes in usage.
Figure 2: Horizon on PowerEdge R7525 solution user density figures
Conclusion
Our performance testing achieved excellent consolidation ratios for the solution while maintaining good performance for typical VDI workloads. PowerEdge R7525 servers based on AMD processors come with dual-socket CPUs that can host up to 128 cores per server, increasing user density within VDI environment.
If you are running a mixed workload on your hypervisor, including your VDI workload, there is a limitation using VMware licensing greater than 32 cores. See the licensing details here. However, this limitation doesn't apply to VMware vSphere Desktop edition intended only for deploying desktop virtualization and is licensed based on powered-on desktop virtual machines.
The high CPU core density per server results in exceptional user densities and high performance for VDI workloads. The 2nd Gen AMD EPYC processors with high core counts present an opportunity to design your VDI environment with CPU oversubscription ratios that result in the right balance between performance and user density.
Related Blog Posts
VDI Data Protection - Part 3: An Operational Backup Approach for Horizon 7
Mon, 12 Dec 2022 21:26:48 -0000
|Read Time: 0 minutes
In Part 1 of this blog series we discussed how disaster recovery and operational backup are two significant aspects of Virtual Desktop Infrastructure (VDI) data protection. In this blog, we will discuss the operational backup aspects of VMware Horizon data protection. For details on disaster recovery, see Part 2.
Loss of VDI environment availability or data has the potential to degrade a user’s ability to perform daily operational tasks. So, it is important for organizations to have an optimal plan to back up and recover VDI data. A robust data protection plan should meet the availability, Recovery Time Objective (RTO), and Recovery Point Objective (RPO) targets defined in Service Level Agreements (SLAs).
For a VMware Horizon virtual desktop environment, three key component layers require protection:
- The desktop layer, that is, the user’s desktop (which is often made available to multiple users using an appropriate provisioning technology)
- The management layer (which performs the provisioning, brokering, policy management, and related management functions)
- The user data layer (stored in user profile shares, home folders, and so on)
The backup and recovery requirements of each component layer depend on the type of the desktop pools and provisioning method used in the Horizon 7 environment. For example, a persistent (stateful) desktop pool can be created with full clones or full virtual machines, which requires a full backup of the virtual machines. A persistent pool can also be created with Horizon instant clones or linked clones with App Volumes (App Stacks and User Writable Volumes) to store the user-installed apps and user-related data. In this scenario, the gold image of the desktop and the persistent data related to App Volumes need protection.
For a non-persistent (stateless) desktop pool, only the gold image of the desktop needs to be protected. In the case of non-persistent desktops, you should consider protecting the user data that is stored in user profile shares and home folders, based on the user environment.
Figure 1: Horizon 7 Operational Backup Approach
Dell EMC offers comprehensive backup and recovery solutions that include products like Integrated Data Protection Appliances (IDPA), Avamar, Data Domain, and Data Protection Suite. For the data protection of a Horizon 7 environment, you can choose from this broad range of Dell EMC data protection products to match your user environment and existing data protection regime. For further information, visit the Dell EMC Data Protection web page.
The Dell EMC Ready Solutions for VDI team has published an operations guide that outlines how Avamar Virtual Edition (AVE) and Data Domain Virtual Edition (DD VE) can be used to facilitate backup and recovery of a Horizon 7, non-persistent desktop pool provisioned by instant clone technology. AVE and DD VE are the software-defined versions of the industry-leading Dell EMC data protection products Avamar and Data Domain. Avamar facilitates fast and efficient backup and recovery for a Horizon environment. Variable-length data deduplication, a key feature of Avamar data protection software, reduces network traffic significantly and provides better storage efficiency. Data Domain provides backup as well as archival capabilities. Data Domain’s tight integration with Avamar delivers added performance and scalability advantages for large Horizon 7 environments. Let’s see some of the key points discussed in the operation guide for backup and recovery of Horizon 7 desktop, management, and user layers.
The Horizon 7 configuration details are in the management layer stored in a View LDAP repository as part of the connection server configurations details. To schedule backups of this database, select the connection server instance from the Horizon console to generate a configuration backup file in a file share. You can then use Avamar VE to back up and restore this configuration backup file. If you are using linked clones, you also need to back up the Composer database.
As discussed earlier in this blog, the backup requirements of the desktop layer depend on the desktop pools and provisioning method. In the case of Horizon instant clones, only the gold image of the respective desktop pools need to be backed up. We recommend taking a clone of the original gold image (containing snapshots) and use that copied cloned image for the backup cycles.
The user data layer contains user-profile shares and other user-related files that are backed up by Avamar software. This layer needs to be protected using a standard data protection approach that is appropriate for user data in any environment.
For a more detailed description of the process to protect each of the layers described above, refer to the operations guide published by the Dell EMC Ready Solutions for VDI team.
The backup and recovery approach for Horizon virtual desktop environments is different from the approach followed for physical desktops and other virtual machines. For developing a successful operational backup strategy for Horizon, the key thing to be aware of is that all three component layers (desktop, management, user data) must be considered. The successful recovery of each of these interdependent components is essential to restore and deliver a fully functional user desktop. To make sure that your backup and recovery plan is effective from a user and business perspective, we recommend that you perform a backup and recovery test for all three layers simultaneously.
In the next part, we will conclude the blog series with some discussion on multi-cloud and hybrid cloud strategies for Horizon 7. So, stay tuned for more!
Thanks for Reading,
Anand Johnson - On Twitter @anandjohns
The Effect of Memory Speed on VDI User Density
Wed, 19 Aug 2020 23:34:48 -0000
|Read Time: 0 minutes
In most modern-day virtual desktop infrastructure (VDI) deployments, RAM (often referred to as memory) is not a bottleneck. More often, it is the processor that gets saturated before memory and storage does. However, it is not recommended to overcommit memory capacity for VDI deployments. It is important that there is a balance between the memory required by virtual machines and host physical memory. Low memory allocation can cause increased storage I/O due to excessive paging. Conversely, if RAM allocation is too high, it affects storage capacity negatively due to the increased size of page files, virtual machine swap files and suspend files.
Memory speed or speed of the memory bus is one of the other attributes of RAM, apart from memory capacity, that may affect the performance of your VDI system. The Dell EMC Ready Solutions for the VDI team recently completed some performance analysis work to check the impact of memory speeds on the ‘density optimized’ configuration offered as a part of Solutions for VDI. The density optimized configuration is based on Intel Xeon Scalable 2nd generation processors code-named Cascade Lake. In this blog, we will discuss the details of this performance analysis work to understand the effects of memory speed on VDI system performance.
Test environment
The VDI Engineering team performed tests with Login VSI, an industry standard tool for benchmarking VDI workloads. The tests were done using Login VSI Knowledge Workload running on VMs configured with 2 vCPUs and 4 GB of RAM with that 4 GB of RAM all being reserved memory.
The testbed environment was a 3-node cluster of VxRail V570F appliances that was optimized for VDI workloads. The cluster was configured and tested with 768 GB of memory per node with a speed of 2666 MHz for test1 and a speed of 2933 MHz for test2. The Environment configuration used was:
- PowerEdge R740xd servers
- Intel Xeon Gold 6248, 2 x 20-core, 2.5 GHz processors
- 768 GB memory (24 x 32 GB @ 2666 MHz) (2 DIMMS per channel (DPC) for test1)
- 768 GB memory (12 x 64 GB @ 2933 MHz) (1 DIMM per channel (DPC) for test2)
- vSAN hybrid data store using an SSD caching tier
- VMware ESXi 6.7 hypervisor
- VMware Horizon 7.7 VDI software layer
The compute workload virtual machines were Windows 10, 64-bit, version 1803. One of the VxRail cluster nodes hosted both management and compute virtual machines. The other two nodes were dedicated to workload compute. Figure 1 shows the main components involved in this work.
Figure 1 Dell EMC VxRail Solutions for VDI Stack Components
Results Analysis
Now let’s check the Login VSI results from the tests done with memory speeds of 2666 MHz and 2933 MHz. Figure 2 shows the comparison graphs of the Login VSI Index Average values (the average response time for the system). From the graphs, we can see that the difference in response times from the two tests was marginal while sessions were loaded. We can ignore these marginal differences when doing a Login VSI test that is based on random workloads. While reaching a CPU utilization threshold of approximately 85%, active session count was 480 from both tests, implying that memory speed doesn’t affect user densities significantly in a VxRail density optimized configuration based on Intel Xeon Scalable Gold 6248 processors.
Note that the Dell EMC Ready Solutions for VDI team considers 85% of CPU utilization as a threshold because testing and loading the system beyond this value might have a negative impact on the performance and end-user experience. So, in these tests, the system was not stressed to the point of reaching a Login VSIMax. VSIMax shows the number of sessions that can be active on a system before the system is saturated.
Figure 2 Login VSI response time comparison with different memory speeds 2666 MHz vs. 2933 MHz
Login VSI test results metrics are summarized in Table 1 below.
Table 1 Login VSI Test Summary
Figure 3 shows the comparison of processor utilization in tests done with memory speeds of 2666 MHz and 2933 MHz. As shown in the figure, we couldn’t see a notable difference in the processor utilization in these tests. CPU utilization steadily increased during the login phase in both tests. The test with 2933 MHz showed a comparatively lower utilization, however, the difference was marginal. The difference in steady-state average CPU utilization was around 4% in these tests.
Figure 3 Comparison of CPU utilization with 2666 MHz and 2933 MHz memory speed
Summary
To summarize, our tests showed that in a VDI system based on the Dell EMC VxRail Density Optimized configuration powered by Intel Xeon Scalable Gold 6248 processors, an increase in memory speed did not improve the overall performance of the selected application workload significantly. It was also evident from our testing that memory was never a bottleneck during the testing. We did not test with other processor models. The results might vary when tested with other models.
In the next blog, we’ll discuss the effect of different Microsoft Windows operating systems versions on VDI user density. So, stay tuned!