Skip to content

AVD · /// Azure Virtual Desktop Load Testing

Right-size AVD with measured concurrency. Not guesses about Azure cost.

LoadGen ships a 7-step AVD wizard with native ARM discovery — Subscription, Resource Group, and Host Pool surface inline. FSLogix profile loads, connection-broker auth, and per-region scaling captured on the same scenario shape that already works for Citrix.

Validate AVD scaleView pricing

ARM-native discoveryFSLogix-aware€1,099 / week · 50 → 25k vUsers

7-step AVD wizard — Subscription · Resource Group · Host Pool inline.

The Problem

Azure invoices arrive monthly. AVD guesses don’t.

AVD adoption is often built on hand-wave concurrency models. Pilot pools pass; production pools don’t. The components that actually break first hide behind a single ARM resource ID.

Spend forecasts built on guesses.

Without measured concurrency on the actual host pool, every Azure spend forecast is a hand-wave. Compute SKU + RAM + FSLogix share size become moving targets — and Azure invoices line up with reality, not slide decks.

Pilot pools that lie.

A 30-user pilot proves nothing about a 1,500-user finance-close. Without parity testing across host pools — same scenario shape, same workload, same auth — regression surfaces on cutover weekend.

FSLogix + broker behind one URL.

Generic load tools see ‘AVD’. They don’t see the resource group, host pool, FSLogix profile share, or connection broker that actually breaks first under measured load.

Why LoadGen for AVD

Native to Azure. Honest about cost.

The wizard speaks ARM. The agents are VDI agents. The scenario shape carries over from any Citrix or Horizon baseline — so a single PoC produces a real before / after on Azure.

ARM-native discovery

The wizard reads your Subscription, Resource Group, and Host Pool inline. No CLI scripts. No JSON spelunking. Configuration captured once, replayed everywhere.

Same scenario, both stacks

The Citrix .lgs workload that already runs in your environment runs unmodified on AVD. Compare HDX p95 vs ARM-broker latency vs Horizon connection-broker on the same execution.

Honest cost numbers

FSLogix profile-load time, login latency, and AVD-specific p95 captured alongside Azure SKU and host-pool density — so Finance sees what they’re actually buying.

Real-time orchestration

AVD pools, live, under measured load.

The cockpit you operate while a 1,500-user run is in flight on your AVD subscription. Per-region concurrency. FSLogix profile-load time. Per-step latency. All on one timeline.

  • vUsers ramp across warm-up / steady / spike / cool-down — visible as the run progresses.
  • VDI agents allocated per pool, per region — actual session-host pressure, not thread abstractions.
  • Per-step latency + AVD-broker round-trip + login time bound to every scenario.
  • Cockpit data flows into SLA reporting after cutover — same engine, monitoring mode.

Live orchestration cockpit during an AVD run.

Capability

What ships with AVD load testing.

7-step AVD wizard

ARM-native Subscription / Resource Group / Host Pool discovery. Pool type, session host SKU, FSLogix share, connection-broker auth — captured inline.

Learn more

VDI agents for AVD

Activate, Reset, and Kill operations land on every session host. Not thread-based abstractions inside one process — actual session-host pressure measured.

Learn more

SUT Monitoring + Azure binding

FSLogix profile-load, AAD auth latency, and connection-broker counters bound to every test. Azure App Service status alongside session-host availability — same timeline.

Learn more

Live cockpit + per-region testing

Per-step latency, per-region concurrency, FSLogix profile-load time, login p95 — visible as the test runs. Multiple agents per region for honest geo-distributed load.

Learn more

Analytics & AI

AI-flagged login-latency anomalies and Citrix-vs-AVD parity overlays. Surfaces regressions on AVD before users report them on cutover weekend.

Learn more

Citrix → AVD comparison overlay

Same workload, both stacks. Compare HDX p95, login latency, host-pool density, and FSLogix performance side-by-side. The before / after a migration program actually needs.

Learn more

Cross-environment validation

Run Citrix and AVD on the same scenario shape.

Stack-honest comparison — HDX p95 against AVD-broker latency, login time against login time, host-pool density against session-host density. The before-and-after every migration program actually needs.

  • Same .lgs scenario authored once, runs unmodified on Citrix + AVD.
  • Up to 5 runs overlaid on one chart — drill into Moments, Errors, per-step measurements.
  • Protocol-honest measurement: HDX channel pressure on Citrix, AVD-broker latency on AVD.
  • The same data drives wave-gated cutover sign-off — see the Migration LP.

Multi-test overlay — 5 runs side-by-side.

Outcomes

Measured before / after on real AVD subscriptions.

Time to first AVD test

Before

7 days

After

4 hrs

−97 %
Host pool right-sizing accuracy

Before

±40 %

After

±5 %

−87 %
Sessions per AVD host

Before

Guess

After

Measured

real
Cutover-week incidents

Before

8

After

1

−87 %

See it in action

Three operational surfaces for AVD.

AVD wizard

ARM-native discovery + FSLogix capture.

Live orchestration

vUsers ramp + per-region concurrency.

Migration overlay

Wave-by-wave Citrix-vs-AVD baseline.

What AVD teams do with this

Three AVD-load testing windows.

Pilot → production scaling

Stress the AVD host pool against a measured concurrency model derived from the 30-user pilot. Right-size session-host SKU, FSLogix share, and host-pool count with evidence.

See use case

Citrix → AVD migration validation

Run the same scenario shape on Citrix and on AVD. Compare HDX vs AVD-broker latency, login time, host-pool density. Sign cutover with data.

See use case

SLA proof for AVD

Wire AVD load testing into the release pipeline. Block a deploy on login-time regression, FSLogix profile-load drift, or host-pool error rate — before production users notice.

See use case

Validate AVD scale on your own subscription.

We’ll spin up the ARM-discovery wizard against your subscription, fire a representative workload from VDI agents, and show you AVD p95, FSLogix profile-load, and host-pool density — live.

Validate AVD scaleTalk to an engineer

Questions

Frequently asked.

Does the AVD wizard work with personal AND pooled host pools?

Yes. The wizard reads the pool type from ARM and configures the workload accordingly — pooled host pools run multi-session scenarios, personal pools run single-session scenarios. FSLogix configuration is captured the same way regardless of pool type.

Can LoadGen test AVD across multiple Azure regions?

Yes. VDI agents can be deployed in any Azure region (or on-prem). The same .lgs scenario fires from every agent on schedule; the cockpit shows per-region concurrency, latency, and host-pool density side-by-side.

What Azure-specific signals does LoadGen capture?

FSLogix profile-load time, AAD authentication latency, connection-broker latency, and per-host-pool density — bound via SUT Monitoring to every test execution. Azure App Service status checks and Azure Function uptime are wired alongside via the Uptime module.

Can the same workload run on both Citrix and AVD for comparison?

Yes — that’s the migration-testing primary use case. The same .lgs scenario authored once runs unmodified on both stacks. The cockpit overlays HDX p95 against AVD-broker latency for direct comparison, which is what most cutover decisions actually need.

Does LoadGen support Windows 365 / Cloud PC?

W365 / Cloud PC is on a divergent product roadmap from AVD even though they share session-host infrastructure. Several W365-adjacent capabilities work today via the same wizard; specifics depend on your Cloud PC license + image strategy — happy to talk through them on a call.

What does AVD load testing cost?

€1,099 per week at the 50-vUser tier, scaling to 25,000 vUsers. Terms run from 1 week to 5 years. Pricing is published in full on the pricing page — no quote-on-request gate for the Load Testing module.

LoadGen Official Logo