Generic listeners are in the beta phase of development and may not be available on your enrollment. Functionality may change during active development. Contact Palantir Support to request access to generic listeners.
Generic listeners receive raw transport-layer connections, rather than HTTPS requests, WebSocket messages, or emails. Use a generic listener when a system in your private network communicates over a TCP or UDP protocol that other listener types do not support.
Connections reach a generic listener through a secure tunnel. Systems in your network connect to a secure tunnel initiator, which forwards the traffic to the platform. Each generic listener is backed by a compute module where you host a custom server to process the traffic.
This page explains how to create a generic listener, request an inbound connection policy, and configure the listener's compute module. To learn how traffic reaches the platform, how to restrict which systems can connect, and how to troubleshoot connections, review Configure secure tunnel ingress.
| Concept | Description |
|---|---|
| Generic listener | A listener that receives TCP connections or UDP traffic and delivers it to its backing compute module. |
| Inbound connection policy | Authorizes a single generic listener to receive traffic of one transport protocol through one secure tunnel cluster. A listener cannot receive traffic until it has an approved inbound connection policy. |
| Compute module | Created automatically with the listener. Your compute module runs the server that receives the listener's traffic. |
A single listener can own multiple inbound connection policies. For example, a listener can own one TCP policy and one UDP policy, so that the same compute module receives both types of traffic.
Before you create a generic listener, ensure the following:
Generic listeners have the following limitations:
Navigate to Data Connection > Listeners.
Select Create new listener, then select Generic under Generic Listeners.

Choose a name and location for the listener. A compute module is created in the same location.
On the Configure connection step, request an inbound connection policy for the listener.
On the Configure server step, specify which exportable markings the listener allows.
Complete the remaining steps of the wizard to create the listener.
You must have permission to edit a generic listener to request an inbound connection policy for it.
The request is sent for approval. Select View request to follow its status in the Approvals inbox.
The platform allocates two ports for each policy when you submit the request:
| Port | Description |
|---|---|
| Gateway port | The port that clients connect to on each initiator in the cluster. Gateway ports are allocated from the range 15000 to 16000 and are unique for each protocol within a cluster. |
| Upstream port | The port that your compute module receives traffic on. Upstream ports are allocated starting from 10000 and are unique for each protocol within a listener. |
Both ports are displayed on the policy after it is created. A policy's protocol, cluster, and ports cannot be changed after the policy is created.
Each inbound connection policy has its own roles, which are separate from the roles on the listener. On the Permissions step of the request, grant roles to the users and groups that need to view or manage the policy. These permissions also control who can view the policy's approval request.
Inbound connection policy requests appear in Control Panel's Approvals inbox. A policy does not take effect until its request is approved. Once the policy is approved and the listener's compute module is deployed, the initiators in the cluster begin listening on the gateway port within a few minutes. Approval alone does not open the gateway port.
Approving inbound connection policy requests is granted through the Administrate inbound connection policies workflow. By default, this workflow is granted to the Information Security Officer role.
When reviewing a request, verify the following:
A listener's inbound connection policies are displayed in the Inbound connection section of its configuration. Each policy shows its protocol, gateway and upstream ports, and approval status.

Administrators can also view all inbound connection policies that use a cluster. Navigate to the Network egress section of Control Panel, open the Secure tunnels tab, and select the cluster.
Your compute module must run a server that binds to 0.0.0.0 on the upstream port of each inbound connection policy owned by the listener:
On the listener's Configure server page, select Edit in Compute modules to open the compute module. Then configure your container and start the module. Review the compute modules documentation for more information on building and deploying containers.
When a generic listener processes data, marking validation acts as a security control that prevents unauthorized export of marked data. The system ensures that any data consumed by the listener's compute module meets the listener's configured exportable marking requirements.
By default, only data without security markings can be read and incorporated into your compute module inputs. If your compute module needs to process data that carries security markings, you must explicitly configure which markings are permitted for export on the listener's Configure server page. Only a user with the ability to declassify those markings can add them to the configuration.

Once the policy is approved and the compute module is running, configure the initiator's ingress allowlist and connect a client.