Talk to a specialist

VLSM.io SOLUTIONS

Random IPv4 Proxy

The same public ingress IP and port. A different public outbound IPv4 for each new request.

IPv4-only egressConnectivity and intelligence
Globe with IPv4-only outbound connections
Outbound IPv4 • per request203.0.113.24Simulated request #1Illustrative addresses
Estimated reading time: 7 min7 sections
01

Fixed ingress. Different egress for every request.

With VLSM.io Random IPv4 Proxy, the application always connects to the same public ingress IP and port. For every new request, the service uses a different public outbound IPv4. The endpoint configured in the client therefore stays stable while the source presented to the destination changes between requests. The application does not need to replace the proxy IP or port to trigger that change. The service controls this behavior according to the selected mode and the integration established with the client software in use.

This model suits independent queries that can receive a response without requiring the same source on the next call. Authorized workers can share the ingress configuration within the project’s credential and capacity limits. That does not mean sharing business state: cookies, tokens, permissions, and results remain the application’s responsibility. If a sequence requires the same outbound IPv4 across steps, compare that requirement with static mode or a suitable rotating window. The random mode described here changes egress per request while preserving the configured ingress endpoint.

INGRESS AND EGRESSPublic IP:PortPublic outbound IPv4
ApplicationConfigures IP:Port
Public ingress198.51.100.10:8080
Public IPv4 egress203.0.113.24
The application keeps the same public ingress IP and port. Static mode retains egress; rotating mode changes public outbound IPv4 at the configured interval in minutes; random mode uses a different egress address for every new request. Addresses and port are illustrative examples.
02

What a new request means

The phrase “for each new request” must be interpreted within the contracted protocol and integration. An application may send several HTTP calls over one connection, keep an HTTPS tunnel open, or multiplex operations. In that situation, the intermediary cannot necessarily choose a new egress for every internal call. Validate the policy at the point where the service actually receives an eligible selection unit. The client may need to create a new connection or use a specific service interface to obtain the intended behavior.

Acceptance testing should therefore use the production library rather than only a different testing utility. Check whether connections are reused, how retries work, and when the connection pool is refreshed. Confirmation should observe the actual egress at a controlled destination. Different addresses demonstrate selection at different moments; they do not alone establish that every application operation followed the same rule. Associate each observation with its task identifier, and retain enough context to explain discrepancies between a simple test and the production workload.

During validation, keep the same public IP and port for every test call. Send consecutive requests through the agreed integration path and record the outbound IPv4 observed for each one. A new request must present a different egress address from the previous request. If it does not, investigate connection reuse, client configuration, and the mode actually applied before increasing traffic. Manually changing the endpoint between calls would invalidate the demonstration that the service itself performs the egress change while the client retains the same ingress configuration.

03

Per-request changes and egress distribution

A different egress address for every new request does not mean an address can never be used again later. The required difference is between the current and previous request; permanent uniqueness across the entire history would be an additional condition. The address pool remains a resource sized for the project. Assess distribution over time and document how operations handle addresses temporarily removed from the pool. This distinction allows teams to measure the promised rotation without assuming an unlimited supply of previously unused public addresses for every future operation.

Monitoring distribution over a representative period helps explain operations. Observe completed tasks per egress, error concentration, and the destinations involved while retaining only the information needed for troubleshooting. Do not turn small samples into conclusions about the entire pool. A concentrated failure may come from a particular destination, incorrect configuration, or a network issue. Analysis should combine source, time, and result to guide a useful action. Comparing these observations with the requested workload is more informative than counting distinct IP addresses alone.

TRAFFIC AND ADDRESSINGSame ingress. New request, different egress.
Fixed ingressOutbound IPv4
Request 1198.51.100.10:8080203.0.113.24
Request 2198.51.100.10:8080203.0.113.78
Request 3198.51.100.10:8080203.0.113.109
The application always uses 198.51.100.10, port 8080. For every new request, the proxy uses a public outbound IPv4 different from the previous request. The change occurs at egress, without changing the IP or port configured in the client. Illustrative addresses and port.
04

Queues, isolation, and consistent results

A queue separates task creation from query execution. Each item should contain the necessary identifier, a known state, and a completion criterion. A worker retrieves the item, establishes the appropriate connection, and records the result. If the operation fails, the process decides whether a retry or review is appropriate. This design is especially important when egresses vary, because it avoids making work continuity depend on a network identity that does not remain fixed. Application state becomes explicit and independent of source selection.

Keep cookies and credentials separate when tasks belong to different accounts or customers. A new source does not create new authorization or make session-data sharing safe. For operations that modify information, define a way to recognize already confirmed results before repeating a call. When a destination provides an appropriate API, evaluate that integration directly. The proxy handles connectivity; consistency, permissions, and result organization remain application responsibilities. Make those boundaries visible in the runbook so developers do not treat an IP change as a substitute for an application decision.

05

The protocol defines the selection boundary

A random integration needs to know where selection occurs. With HTTP, compare behavior observed in individually forwarded requests with connections reused by the client. With HTTPS through CONNECT, the tunnel carries the encrypted session and does not necessarily represent a new egress choice for each internal call. A library may maintain a connection pool even when the application creates separate tasks. Inspect that configuration and validate results with a controlled destination before claiming that every task uses an independent IPv4 selection within the contracted service.

With SOCKS5, the destination hostname, connection, and any UDP association also need to be understood as integration elements. SOCKS5H is the proxy-side resolution option in clients that provide it, without creating a new protocol or a guarantee of randomness. UDP support must be explicitly confirmed and cannot be inferred from TCP support. Map each tool to its actual behavior: a browser, query library, and collection agent may share the same endpoint while producing different selection boundaries because of how they establish and reuse connections.

06

Compare random mode with the alternatives

Static mode keeps the same public outbound IPv4. Rotating mode keeps the public ingress IP and port but changes the outbound IPv4 at the previously configured interval in minutes. Random mode also preserves ingress and uses a different egress address for every new request.

The distinction between rotating and random is therefore the event that triggers the change: time or request. Choose a mode based on task duration, continuity requirements, and destination behavior, keeping access configuration separate from the egress policy applied by the service.

This solution uses IPv4 egress. An IPv6-only destination requires different connectivity; having IPv6 on the local network does not change that requirement. Tell VLSM.io about the destinations, tools, expected volume, and session rules. Define acceptance testing with criteria for egress selection, task completion, and failure handling. The agreement can then reflect verifiable behavior, and the team gains an objective reference for monitoring future application or service changes. Keep the tested client configuration alongside the operational documentation so later upgrades can be compared with that baseline.

FROM CONNECTION TO RESULTTurn requirements into an observable operation.
Plan
Connect
Monitor
Destination planning, connectivity configuration, and result validation are complementary steps. The animation represents this workflow.
07

Assess distribution and completed work

Monitor two conditions separately: stability of the ingress IP:port and the outbound IPv4 change for each new request. Compare consecutive calls and record any deviation from the defined behavior. Then assess pool distribution over a representative period without confusing later reuse of an address with a failure to rotate between consecutive calls.

Associate results with task type and destination. Correct completion of useful work remains the most valuable business indicator, together with evidence that the connection policy is being fulfilled under the application’s actual operating conditions.

Independent tasks can create bursts of new connections, with TLS negotiation cost and resource usage in every worker. Size those effects alongside bandwidth and concurrency rather than using only the number of IPs as a reference. Network offerings from 1 Mbps to multiple 100 Gbps depend on the project and do not represent the individual capacity of each egress address. Set queue limits, reduce attempts when the destination imposes throttling, and preserve processing capacity for handling responses and storing the data that the application actually needs.