Windows access control
Browser RDP access for Windows servers, scoped by CertLocker tokens
RDP is still how a lot of real infrastructure gets managed. CertLocker now lets teams bring Windows desktop access into the same short-lived, scoped, auditable workflow already used for SSH.
A lot of privileged access conversations quietly skip Windows desktop access. They focus on SSH keys, Kubernetes exec, cloud consoles, and identity providers, while the systems people actually support still include RDP into Windows servers, IIS boxes, store systems, warehouse tools, legacy admin apps, and supplier-supported machines.
The awkward part is rarely the RDP protocol itself. The awkward part is everything around it: who is allowed to open the desktop, which host they can reach, which credential is used, how long access lasts, and what evidence remains after the session.
CertLocker now treats RDP as a first-class protocol on access tokens. A token can be short-lived, bound to groups, limited to selected hosts, attached to the right credential secret, and launched from the browser when the operator needs the desktop.
Why saved RDP files are not an access model
The common pattern is familiar: a VPN, an internal hostname, a shared admin password, and a saved .rdp file. It works, which is why it survives. It also tends to leave poor answers to simple operational questions.
- Who still has a copy of the connection file?
- Which credential is inside the workflow, and when was it rotated?
- Does access expire automatically after the job?
- Can support reach one Windows host without inheriting broad network access?
- Can the team show when the access path was created, changed, suspended, or used?
CertLocker moves those decisions into the same place teams already manage certificates, secrets, SSH access, probes, groups, and audit records. The desktop becomes an access path governed by a token, not a loose file that keeps working until someone remembers to clean it up.
The token defines the desktop access window
In the token editor, RDP is selected explicitly as a protocol. That matters because desktop access should not be implied just because a token exists. A token can allow SSH, SFTP, SCP, RDP, or a narrower subset depending on the job.
The useful control point is the token lifecycle. The access grant can have a duration, an absolute expiry, single-use behavior, auto-rotation behavior, group scope, and protocol scope. If the token expires, is revoked, or no longer matches the requested host or protocol, the RDP path should not open.
Hosts remain explicit
RDP access is tied to a host record, not an operator's memory of an IP address. That keeps routing visible and reviewable. The token detail view shows which Windows machine is in scope before anyone launches the desktop.
That is especially useful for mixed estates: store Windows machines, warehouse systems, back-office IIS servers, point-of-sale support hosts, and customer environment jump boxes. The access request should say which host is being reached, not just which network someone was allowed to enter.
The browser session is the convenience layer, not the security model
Browser RDP is convenient because an operator can launch the desktop without juggling a separate client, copied connection details, or a local file. But the browser is not the core trust decision. The trust decision is the token and the policies around it.
CertLocker checks the token, host scope, protocol scope, credential binding, user context, and current token state before the desktop path becomes usable. The browser session is where the operator works after those checks pass.
This keeps Windows access aligned with the rest of the infrastructure trust cycle:
- Define ownership and groups.
- Create a short-lived access token for the job.
- Bind the token to the correct Windows host and credential secret.
- Enable RDP only when desktop access is required.
- Launch from the browser after the access decision passes.
- Keep the token, host, credential, and session activity inside the audit story.
Where this helps most
This is not aimed at replacing every remote desktop tool. It is aimed at the operational gap where RDP is necessary but the access process around it is too loose.
Retail infrastructure is a good example. Stores, warehouses, supplier portals, payment-adjacent systems, and old Windows applications often sit outside the clean cloud-native story. They still need controlled access, sensible expiry, credential governance, and evidence when someone asks who opened what.
The same applies to MSP handoffs, vendor support windows, customer-hosted deployments, Windows jump boxes, and IIS estates that still need occasional desktop access.
What to test
If you want to evaluate this properly, do not test it against a perfect demo host. Test it against the Windows systems your team actually operates: the awkward one in the branch, the IIS server with manual steps, the support box a supplier sometimes needs, or the jump host everyone knows is important but nobody enjoys owning.
The hosted evaluation at trust.certlocker.io is the quickest way to walk through the workflow. For a real technical validation, run a one-month onsite evaluation and connect CertLocker to the Windows hosts, groups, and access paths in your own environment.