Agent-based discovery: Benefits and use cases | JDisc
Introduction
A server can be running perfectly and still be absent from your inventory. Your discovery server may have no permission to connect into its network segment. A laptop may rarely be reachable from the office. In both cases, repeating the same remote scan leaves the same gap.
Agent-based discovery collects inventory on the computer itself through installed discovery software. The computer can then initiate the transfer to a central inventory system. This is useful when the system can send data out but cannot accept the remote-management connections that a central scan needs.
JDisc Discovery supports this approach with the JDisc Discovery Local Scanner for Windows. It complements remote discovery with local collection, scheduled scans, immediate scans, HTTPS uploads, and scan-data export. Here we explain when that approach helps, what you need to plan, and how to combine it with agentless discovery.
What is agent-based discovery?
Agent-based discovery is an inventory method in which software installed on a target computer collects information locally. Rather than having a central server query that computer through remote-management interfaces, the local scanner performs the collection on the machine being inventoried.
Collection and delivery are separate operations. A local scan produces inventory data; an upload transports that data; the receiving discovery installation imports it. A successful installation therefore does not prove that current inventory is available centrally.
The term agent-based scanning describes the same local collection model. It does not automatically mean continuous monitoring, vulnerability assessment, remote control, or patch deployment. Those are separate capabilities that must be evaluated individually. On this page, the focus is inventory collection with Local Scanner.
The scanner inventories its own Windows computer. Installing it on one server does not make that server a discovery proxy for all other devices in its subnet.
How agent-based discovery changes the connection model
In a remote scan, the discovery system needs a permitted path to the target’s management interfaces. With Local Scanner, the target collects its own data and initiates an HTTPS connection to the configured Discovery Web UI server.
That change in direction can fit a security policy that blocks incoming management connections but allows a specifically approved upload. It can also avoid making an endpoint reachable through a VPN merely for an inventory scan, provided that endpoint can reach the receiving service through another approved route.
There are still prerequisites: name resolution, routing, the configured HTTPS port, and appropriate certificate validation. Local collection does not bypass a firewall or create connectivity. HTTPS also describes a normal bidirectional network connection; an endpoint-initiated upload is not a physical one-way link.
Where agent-based discovery has advantages
Protected server tiers
A central discovery installation may sit in a management segment that is deliberately prevented from opening sessions into a more sensitive server tier. Opening additional remote-management paths solely for inventory can conflict with that design.
Local collection offers another option: install the scanner on the protected Windows server and permit a defined outbound upload to the receiving installation. You can collect inventory without requiring the discovery server to initiate a remote scan into that server’s tier.
This changes the access pattern, not the need for a security review. The local software, its execution permissions, the receiving service, and the data leaving the protected tier still need to fit your policy.
Remote computers that rarely accept inbound connections
A home-office laptop may be behind a router, a host firewall, or a network policy that prevents central management connections. Even when the computer is in use, it may be unavailable from the discovery server’s perspective.
An endpoint-initiated upload can be easier to accommodate than inbound remote management. For Local Scanner, the important condition is that the Windows computer can reach its configured HTTPS receiver. It does not need to be directly reachable for a central scan at that moment.
Do not assume that an internal receiver is reachable from every home network. Decide how the approved connection will work before deployment. This architecture alone does not establish internet access or guarantee delivery while a computer is offline.
Segments with restricted remote-management protocols
A Windows computer may be reachable on the network while the interfaces needed for detailed remote inventory remain blocked. Basic reachability and usable management access are different things.
If policy permits local discovery software, agent-based scanning can move collection onto the computer instead of requiring that remote-management access. You still have to authorize local execution and the upload path. The advantage is a different set of prerequisites, rather than the absence of prerequisites.
Inventory collection during maintenance or an on-site visit
Sometimes you need an inventory immediately after a hardware change or software rollout. Waiting for a central scan window is inconvenient, especially when an engineer already has access to the computer.
Local Scanner can start an immediate scan from its local interface. Choose scan and upload when the receiver is reachable, or scan and ZIP export when you need a file for a controlled handover. Exporting a file does not update the central database by itself.
Collection without an immediate upload
An isolated computer may have no approved connection to the receiving installation. A local scan with ZIP export lets you separate collection from transfer and subsequent processing.
This is useful for an agreed offline collection workflow, but it does not create automated inventory freshness. Plan who transfers the scan data and how it will be processed. Treat exported inventory files as sensitive operational information and handle them according to your organization’s rules.
The important detail is the blocked connection from Tier1 into Tier0. The Local Scanner on the Tier0 server collects that server’s inventory and initiates an HTTPS upload toward Tier1. The discovery installation receives and imports the data. Tier2 clients can continue to use direct discovery.
A tiering example: Local collection where direct scanning is blocked
The diagram illustrates three zones. Tier0 contains a protected Windows server. Tier1 hosts JDisc Discovery. Tier2 contains client computers that the discovery installation can scan directly.
Tier0, Tier1, and Tier2 are the labels used in this example. The diagram explains the collection paths; it is not a universal tiering policy or an endorsement of every cross-tier connection. Whether a protected system may upload inventory to a particular receiver depends on your own architecture and security requirements.
Also notice the scope: the Tier0 scanner collects from its own server. Other protected Windows servers would need their own local collection arrangements.
Agent-based vs. agentless discovery: Which method fits?
The useful question is which collection method works for each part of the environment. Our agentless network discovery overview explains the remote approach in more detail.
| Situation | Practical starting point |
|---|---|
| A Windows server blocks incoming discovery connections but can upload to an approved receiver | Local Scanner with an endpoint-initiated HTTPS upload |
| A remote Windows computer can reach the receiver but cannot be scanned centrally | Local Scanner, with an appropriate local scan schedule |
| A computer needs inventory collection without immediate network delivery | Local scan with ZIP export and an agreed transfer workflow |
| Reachable computers already expose the required management interfaces | Agentless discovery avoids an additional scanner rollout |
| Switches, routers, printers, or appliances cannot run the Windows scanner | Use supported network protocols or platform APIs |
| You want to find devices that have never received an agent | Include network-based discovery in the coverage plan |
An installed agent only helps on computers where it is deployed and functioning. It cannot reveal every unmanaged device on the network. Conversely, a centrally managed remote scan cannot collect detailed data through a connection that policy does not allow.
Combining the two methods lets you choose the path by device and network conditions while keeping the results in a central network inventory.
How JDisc Local Scanner fits into your inventory
For regular collection, prepare a Windows deployment package in the JDisc Discovery Web UI. The configuration specifies the receiving server and local collection schedule. Distribute the package through your normal software deployment process or install it on selected computers.
When the scanner first contacts the receiving installation, its registration needs approval before inventory uploads are accepted. Verify the computer’s identity before approving it and apply the configuration change. Authorization and successful import are separate checks.
For operational verification, look at the Local Scanner Registrations report. A registration state of Approved shows upload authorization. An Import Status of Completed, together with a sufficiently recent Last Import Time, confirms a recorded successful import. The timestamp must be read with the outcome; it can also record a failed import attempt.
For installation, scheduling, registration approval, immediate scans, and troubleshooting, use our dedicated Local Scanner technical article.
The Local Scanner video walkthrough shows the tiering example and follows deployment, installation, registration approval, and the resulting inventory workflow.
What to plan before deployment
Scanner lifecycle and local permissions
An agent introduces software that needs deployment, updates, and operational ownership. Check installation outcomes and decide who investigates missing results. A computer without a working scanner remains a coverage gap even if it was included in the rollout plan.
Local execution also requires appropriate permissions. Agent-based discovery changes where collection runs; it does not remove privilege requirements. Review those permissions alongside the accounts and access paths used for remote discovery.
Receiver access and certificate validation
Restrict the upload path to the intended receiving service and configured port. Use a server certificate that the target computers can validate. Local Scanner’s option to trust any TLS certificate removes the normal certificate-based assurance of the receiver’s identity; it should not be used as a general fix for connection problems.
The upload design does not imply that your Web UI should simply be exposed to the public internet. Select and review the receiving architecture for the actual network locations involved.
Inventory freshness and scope
Choose scan frequency according to how quickly the systems change and how current the data needs to be. Scheduled inventory is a sequence of observations, not evidence that every change was captured as it happened.
Track successful imports, rather than installation counts alone. Reconcile the intended rollout with registrations and import outcomes. For offline exports, include the collection time and handover process in that review.
The Local Scanner workflow described here is for Windows computers. Broader agent-based product claims, operating-system support, and inventory depth should not be assumed from the word “agent.” Likewise, this page makes no claim that every remote-discovery field or infrastructure relationship is available through local collection.
Start with the collection path your systems allow
Agent-based discovery is particularly useful when a computer can collect inventory locally but cannot accept a central remote scan. Protected server tiers, remote Windows computers, and controlled offline collection are practical examples.
Use that capability where it solves an access problem, and keep agentless scanning for the infrastructure and devices it serves well. The resulting inventory can support IT asset management, reporting, and downstream processes without forcing every system through the same collection method.
Begin with one representative Windows computer. Verify local collection, registration approval, delivery, and a successful import before expanding the rollout. To evaluate JDisc Discovery, start with the trial download.
Frequently Asked Questions
These questions explain when local collection helps and what it changes in the discovery architecture. They also cover the practical limits of Local Scanner, from platform scope to connection requirements and import verification.
Agent-based discovery uses installed software to collect inventory on the target computer. The results can then be transferred to a central inventory system. JDisc Discovery implements this local collection approach for Windows with Local Scanner.
It is useful when incoming remote-management connections are restricted but local execution and an outbound upload are permitted. It also supports immediate local collection and file export when direct delivery is unavailable.
For the upload workflow, the computer initiates HTTPS communication with the configured Discovery Web UI server. A central discovery connection into the target computer is not required for that local scan and upload. The outbound path must still be allowed.
No. It scans the Windows computer on which it runs. It is not a subnet scanner or a discovery proxy for neighboring devices.
Yes. An immediate local scan can export the collected inventory to a ZIP file. That export does not automatically update the central database; transfer and processing remain separate steps.
Security depends on the complete implementation. Local collection can fit a restrictive connection policy, but it introduces software and execution permissions on the target. The receiving service, certificate validation, approval process, and inventory data also need appropriate controls.
The workflow described here provides inventory collection through scheduled or immediate scans. It should not be interpreted as a promise of continuous monitoring, remote control, vulnerability assessment, or patch deployment.
Check the registration and import result together. Approved authorizes uploads; Completed with a recent Last Import Time confirms a recorded successful import. Review the outcome and events because an import timestamp alone can also refer to a failed attempt.
Use both according to your environment. Local Scanner can cover selected Windows computers with restricted remote access. Agentless methods remain useful for reachable systems, infrastructure devices, and discovering computers that have never received an agent.



