Hi AVD Community,

I hope everyone is doing well! I’m reaching out to gather insights and recommendations on optimizing the networking configuration for Azure Virtual Desktop (AVD) environments. Our organization has been running AVD for several months, and while the overall performance is satisfactory, we are looking to fine-tune the networking aspects to ensure maximum efficiency and performance.

Current Setup

AVD Host Pools:We have two host pools: one for general productivity applications (Office 365, browsers) and another for more resource-intensive applications (CAD, data analysis tools).Each host pool consists of multiple session hosts running on D8s_v3 and NV6 VMs respectively.Networking:All session hosts are within a single Virtual Network (VNet) with a 10.0.0.0/16 address space.Subnets are divided as follows:Prod-Subnet: 10.0.1.0/24 for general productivity applications.Resource-Intensive-Subnet: 10.0.2.0/24 for resource-intensive applications.We use an Azure VPN Gateway to connect our on-premises network to the Azure VNet.Network Security Groups (NSGs) are applied to control traffic flow.Routing and Access:Forced tunneling is enabled to route all internet-bound traffic through our on-premises firewall for security monitoring.We have configured custom routes to ensure traffic between subnets and on-premises is optimized.Load Balancing:Azure Load Balancer is used to distribute user sessions across the VMs in each host pool.

Challenges and Questions

Latency and Throughput:We observe varying latency and throughput between session hosts and our on-premises network. What are the best practices to minimize latency and maximize throughput in such a hybrid setup?Network Security Groups (NSGs):Our NSGs are configured with a mix of service tags and IP ranges. How can we optimize NSG rules for better performance without compromising security?VNet Peering and Performance:Would VNet peering between different host pools improve performance, especially for applications that need to communicate frequently between host pools? What are the potential pitfalls?VPN Gateway Optimization:Given that we use forced tunneling through our on-premises VPN gateway, how can we ensure that this does not become a bottleneck? Are there specific VPN gateway configurations that could enhance performance?Load Balancer Configuration:How can we optimize Azure Load Balancer settings for session persistence and faster failover times? Are there specific configuration tweaks that can help?Monitoring and Diagnostics:What tools and metrics should we focus on for monitoring network performance within AVD? How can we proactively diagnose and address network-related performance issues?

Looking for Recommendations

I would greatly appreciate any detailed recommendations, best practices, or experiences you can share regarding the above challenges. Specific configuration examples, performance metrics, and case studies would be extremely helpful.

​Hi AVD Community, I hope everyone is doing well! I’m reaching out to gather insights and recommendations on optimizing the networking configuration for Azure Virtual Desktop (AVD) environments. Our organization has been running AVD for several months, and while the overall performance is satisfactory, we are looking to fine-tune the networking aspects to ensure maximum efficiency and performance. Current SetupAVD Host Pools:We have two host pools: one for general productivity applications (Office 365, browsers) and another for more resource-intensive applications (CAD, data analysis tools).Each host pool consists of multiple session hosts running on D8s_v3 and NV6 VMs respectively.Networking:All session hosts are within a single Virtual Network (VNet) with a 10.0.0.0/16 address space.Subnets are divided as follows:Prod-Subnet: 10.0.1.0/24 for general productivity applications.Resource-Intensive-Subnet: 10.0.2.0/24 for resource-intensive applications.We use an Azure VPN Gateway to connect our on-premises network to the Azure VNet.Network Security Groups (NSGs) are applied to control traffic flow.Routing and Access:Forced tunneling is enabled to route all internet-bound traffic through our on-premises firewall for security monitoring.We have configured custom routes to ensure traffic between subnets and on-premises is optimized.Load Balancing:Azure Load Balancer is used to distribute user sessions across the VMs in each host pool.Challenges and QuestionsLatency and Throughput:We observe varying latency and throughput between session hosts and our on-premises network. What are the best practices to minimize latency and maximize throughput in such a hybrid setup?Network Security Groups (NSGs):Our NSGs are configured with a mix of service tags and IP ranges. How can we optimize NSG rules for better performance without compromising security?VNet Peering and Performance:Would VNet peering between different host pools improve performance, especially for applications that need to communicate frequently between host pools? What are the potential pitfalls?VPN Gateway Optimization:Given that we use forced tunneling through our on-premises VPN gateway, how can we ensure that this does not become a bottleneck? Are there specific VPN gateway configurations that could enhance performance?Load Balancer Configuration:How can we optimize Azure Load Balancer settings for session persistence and faster failover times? Are there specific configuration tweaks that can help?Monitoring and Diagnostics:What tools and metrics should we focus on for monitoring network performance within AVD? How can we proactively diagnose and address network-related performance issues?Looking for RecommendationsI would greatly appreciate any detailed recommendations, best practices, or experiences you can share regarding the above challenges. Specific configuration examples, performance metrics, and case studies would be extremely helpful.  Read More

​