Citrix · /// HDX & ICA Load Testing

Validate Citrix capacity before production — not in the incident review.

LoadGen runs the only platform-specific Citrix load test on the market: a 7-step wizard for Basic (ICA) and Enhanced (HDX), StoreFront / Gateway / PNAgent / External Login authentication, and Citrix-specific .lgs workloads on Full agents. First test in 4 hours, not 5 days.

See it on your Citrix stack Talk to an engineer

4-hour first testFull HDX + ICA coverage€1,099 / week transparent

/New Citrix Profile

Mode: Basic Online

AD

admin

Administrator

New Citrix Profile

Step 1: Set profile name and Citrix type

Citrix Enhanced

Basics

Connection

Users & Agents

Display

Scenario

Advanced

Review

Profile Basics

Configure the basic settings for your Citrix load profile.

Profile Name *

A unique name to identify this profile

Profile Status

Profile is active

Enable or disable this profile

Citrix Type

DocumentationSupportCreate Ticket

7-step Citrix wizard — Basic (ICA) + Enhanced (HDX), all auth flows.

The Problem

Generic load tools don’t speak Citrix.

Citrix is a protocol stack, an authentication topology, and a delivery model — not just an HTTP endpoint. Tools that ignore that produce numbers that lie.

HDX channel saturation gets smoothed away.

HTTP samplers miss ICA round-trip, HDX channel pressure, and StoreFront / Gateway auth latency. Generic tools report green dashboards while real Citrix sessions stall.

Auth flows trip every other tool.

StoreFront, Gateway, PNAgent, and External Login each have their own quirks. Without dedicated handling, capacity tests stall before the user even opens an app.

Capacity claims that don’t survive Monday.

A 50-user smoke test in DEV cannot model 1,500-user finance-close traffic on production session hosts. Generic load shapes break first on the components that hurt most.

Why LoadGen for Citrix

Citrix-native depth. Full-platform breadth.

The same scenario engine that drives load tests also drives end-to-end monitoring and uptime. One platform, one cockpit, three operational modes.

Citrix auth flows, native

StoreFront, Gateway, PNAgent, External Login — captured by the 7-step wizard, not by hand-scripted brittle login automation.

HDX + ICA protocol-level signals

Per-step latency, HDX channel pressure, ICA round-trip, and Citrix-specific p95 — captured natively, not synthesised from HTTP sampling.

Same scenario, multi-stage

The .lgs workload that runs your load test also drives continuous monitoring and Citrix-vs-AVD baseline comparison. No re-authoring.

Live orchestration

See the runtime environment, not a slide of it.

The same cockpit your engineers will operate on cutover weekend. vUsers ramping, Full-agent allocation, scenario queue — all live.

vUsers ramp curve with warm-up / steady / spike / cool-down phases.
Full agent allocation per pool — Core / Full / VDI — visible as the test runs.
Concurrency counter + scenario queue + phase ribbon on one scrub bar.
The same scenario shape powers live monitoring after cutover — one engine, two modes.

Load Profiles › Orchestration

Mode: BasicOnline

A

Load Profile · Workload Orchestration

Ramp-up · steady · spike · cool-down, scheduled across agents and scenarios.

Concurrent users

500peak 540

Current phase

Spike32 min elapsed

Agents allocated

148 · 4 · 2

Scenarios

3/5active / total

Ramp profile

vUsers · 60 min window

Warm-up

Steady

Spike

Cool-down

Agent pools

Core8

Full4

VDI2

Scenario queue

orchestrated

Citrix Finance · daily peak

240 vU

AVD Sales · sprint launch

160 vU

Web Storefront · soak run

100 vU

RDS Engineering · spike test

80 vU

Horizon HR · capacity sweep

60 vU

DocumentationSupportCreate Ticket

Live orchestration cockpit — paused on hover.

Capability

What ships with the Citrix load testing module.

7-step Citrix wizard

Basic (ICA) and Enhanced (HDX) modes; StoreFront / Gateway / PNAgent / External Login; display + scenario + agent strategy in minutes.

Learn more

Full agents for Citrix sessions

Per-session Activate / Reset / Kill operations land on every session host. Not thread-based abstractions.

Learn more

SUT Monitoring on Citrix infra

Bind Citrix Broker, StoreFront, Gateway, and CitrixDB counters to every test. Capacity correlates with infra health on the same timeline.

Learn more

Live cockpit, real-time KPIs

Per-step concurrency, latency, error hotspots, and HDX-channel health — visible as the test runs, not buried in a post-test report.

Learn more

AI-flagged HDX anomalies

Analytics AI scores HDX latency drift, login-time regressions, and per-step bottlenecks per execution. No analyst hours per release.

Learn more

Citrix vs AVD comparison overlay

Same .lgs workload runs on both stacks. Compare HDX p95, login latency, and host-pool density side-by-side on the same scenario shape.

Learn more

Outcomes

Measured before / after on real Citrix farms.

Numbers from the customer base — first-test setup time, auth-flow capture, per-agent session density, and Citrix-specific p95 coverage.

Time to first Citrix test

Before

5 days

After

4 hrs

−97 %

Auth-flow setup hours

Before

12 h

After

40 min

−94 %

Sessions per agent

Before

40

After

120

+200 %

Citrix-specific p95 captured

Before

Partial

After

Full

HDX + ICA

See it in action

Three live surfaces that ship with the platform.

Live cockpit

Per-step concurrency, latency, HDX-channel health.

Testing › Active Test › Run #847

Mode: BasicOnline

A

Active Test · Run #847

Streaming via SignalR · 12m 34s elapsed

Events/s

+4.2%

1,842

Concurrent

+312 vs base

2,847

Success

+0.2%

98.7%

Live Throughput

Events/sErrors/s−10m · −5m · now

Live Event Stream

SignalR · hub://loadgen/run/847

14:23:00.000POST /api/runs/events38msOK

14:22:59.767POST /api/alerts/check31msOK

14:22:59.534GET /api/dashboards/meta82msOK

DocumentationSupportCreate Ticket

Orchestration

vUsers ramp + Full-agent pool allocation + scenario queue.

Load Profiles › Orchestration

Mode: BasicOnline

A

Load Profile · Workload Orchestration

Ramp-up · steady · spike · cool-down, scheduled across agents and scenarios.

Concurrent users

500peak 540

Current phase

Spike32 min elapsed

Agents allocated

148 · 4 · 2

Scenarios

3/5active / total

Ramp profile

vUsers · 60 min window

Warm-up

Steady

Spike

Cool-down

Agent pools

Core8

Full4

VDI2

Scenario queue

orchestrated

Citrix Finance · daily peak

240 vU

AVD Sales · sprint launch

160 vU

Web Storefront · soak run

100 vU

RDS Engineering · spike test

80 vU

Horizon HR · capacity sweep

60 vU

DocumentationSupportCreate Ticket

Run comparison

Up to 5 runs overlaid — Moments, Errors, per-step.

Testing › Results › Compare

Mode: BasicOnline

A

Test Analysis - Run Comparison

Overlay up to 5 runs to see performance deltas.

Compare Runs

·1 selected

Last 7 days

Citrix

Runs

Run #843Baseline

Run #844

Run #845

Run #846

Run #847

Avg Response

280 ms

P95 Latency

420 ms

Error Rate

0.12%

Throughput

1450/s

Response Time - Overlay

ms · 16 samples

Run #843 is the baseline

#843

DocumentationSupportCreate Ticket

Where this lives in the year

Three Citrix load-testing windows.

Peak-season capacity validation

Stress-test the Citrix farm against modelled finance-close, year-end, or open-enrolment traffic. Right-size session hosts with measured concurrency, not vendor-estimate spreadsheets.

See use case

Pre-cutover migration baseline

Baseline Citrix Basic vs Enhanced — or Citrix vs AVD — on the same scenario shape. Promote to production behind a measured validation gate, not a guess.

See use case

Release-gate performance proof

Wire LoadGen into the deploy pipeline. Block releases on HDX p95 regressions, login-time drift, or Citrix-broker error rates — before production users notice.

See use case

Run your first Citrix load test in 4 hours.

Bring your StoreFront URL and a workload sketch. We’ll author the .lgs scenario on a call, fire it from a managed Full agent, and show you HDX p95 — live.

See it on your Citrix stack View pricing

Questions

Frequently asked.

Does LoadGen support Citrix Basic (ICA) and Enhanced (HDX)?

Yes — both modes ship in the 7-step Citrix wizard. The same workload can be replayed across Basic and Enhanced, and the analytics surface compares HDX p95 and login latency against the ICA baseline on the same timeline.

How are StoreFront, Gateway, and PNAgent authentication handled?

Each auth flow is natively detected and configured by the wizard. External Login is included. The wizard captures the connection profile once; from then on the auth flow is replayed every test without hand-scripted automation.

What’s the difference between a Full agent and a thread-based simulator?

A Full agent runs as an actual Citrix client session — Activate, Reset, and Kill operations land per-session on the real session host. Thread-based simulators abstract these as logical threads inside one process, missing per-session resource pressure on the host.

Can I correlate test load with Citrix infrastructure counters?

Yes — SUT Monitoring binds Citrix Broker, StoreFront, Gateway, and CitrixDB counters (plus any custom Windows / SQL counters you wire) to every test execution. Infrastructure health and measured load share one timeline.

How long is the first Citrix load test typically?

The published target is 4 hours from intake to first test result: 40 minutes to capture the auth flow, ~1 hour to author the .lgs workload, the rest to schedule and run a representative workload. Compared against the 5-day timeline typical of generic load tools.

What does Citrix load testing cost?

The Load Testing module is €1,099 / week at the 50-vUser tier, scaling to 25,000 vUsers. Terms run from 1 week to 5 years. See the pricing page for tier details.