Application monitoring measures whether the applications and business transactions people depend on are available, responsive, and usable.
Direct answer: Application monitoring is the continuous measurement of application availability, response time, errors, and user transactions. The most useful view starts where the user starts: submit, click, or key to the expected screen-ready state. CitraTest APM provides that user-centric application monitoring for every Windows application, including thick clients and published apps.
What application monitoring should see
Infrastructure dashboards, server counters, code traces, and protocol telemetry are valuable views, but they do not always prove that a person can complete the next step. A green host or HTTP 200 can coexist with a slow thick-client form, a spinning published app, or a chart that has not painted.
- Availability: can the user log in, launch the app, and complete the transaction?
- Response time: how long from the action until the usable result is visible?
- Real user path: does monitoring cover identity, Citrix/RDS/cloud session, app launch, and in-session work?
- Evidence: are timings, screenshots, failures, and p95/p99 trends available when a ticket arrives?
Application monitoring beyond code agents
Many APM tools instrument code, services, and infrastructure. That is essential for finding root cause in software you own. User-centric, on-the-glass APM complements those tools by measuring pixels, dialogs, and ready states that can be hidden inside Citrix HDX, Microsoft RDS/RDP, AVD, or another cloud VDI session.
| Monitoring view | Best at | What it may miss |
|---|
| Host and infrastructure | CPU, memory, disk, profile, network | Whether the business screen is usable |
| Code / full-stack APM | Traces, logs, services, dependencies | Published GUI and thick-client paint |
| Protocol / portal checks | ICA/HDX, RDP, StoreFront, Gateway, HTTP | In-session app readiness |
| On-the-glass application monitoring | What the user sees and can complete | Needs stable screen-ready baselines |
How CitraTest APM monitors applications
- Start a synthetic client from the location and connection path users actually use.
- Drive keyboard and mouse through login, launch, search, save, or another business transaction.
- Use image recognition and OCR to confirm the expected screen state, then record click-to-ready time.
- Alert on availability, response-time budgets, failures, and trends; retain screenshots for diagnosis.
This works for browser and Windows applications, SAP, Salesforce, Epic, Cerner, and published apps delivered through Citrix and RDS. No monitoring agent is installed on production VDAs, RDSH, or application servers.
Monitoring versus load testing
CitraTest APM watches production continuously. CitraTest VU uses the same real-user approach to test capacity and change under concurrency. For the method, see the on-the-glass APM blog guide, the Citrix stack response-time method, and the Citrix load-testing guide.
FAQ
What is application monitoring?
It measures availability and the response time of the application transactions users need, including the visible ready state.
Does it cover thick-client, Citrix, and RDS applications?
Yes. User-centric monitoring can drive the real client and measure the screen, including a published app inside a remote session.
Is application monitoring the same as observability?
No. Observability explains services and infrastructure; application monitoring confirms the user-facing application works. The views work best together.
CitraTest APM® is Tevron's application monitoring solution. On-the-glass describes the user-centric measurement approach.