Talk to a specialist

VLSM.io SOLUTIONS

Random IPv6 Proxy

The same public IPv4 and ingress port. A different public outbound IPv6 for every new request.

IPv4:Port ingressIPv6-only egressConnectivity and intelligence
Globe with outbound IPv6 connections; proxy ingress over IPv4 and port
Outbound IPv6 • per request2001:db8::24IPv4 ingress: 198.51.100.10:8080Simulated request #1Illustrative addresses
Estimated reading time: 7 min7 sections
01

IPv6 distribution for independent tasks

With VLSM.io, the application connects to the proxy’s public IPv4 and ingress port, while the destination-facing connection uses a public outbound IPv6. The Random IPv6 Proxy serves projects in which different tasks can use different outbound addresses without depending on a permanent network identity.

An independent public information request, an availability check, or an authorized catalog update are examples of activities that can be evaluated with this model. Random selection takes place within the pool made available to the project. Its purpose is to organize application connectivity with capacity planning and monitoring, without turning the number of available addresses into a promise of unrestricted access.

Before selecting the service, identify what constitutes a complete task. Reading an isolated page differs from signing in, querying several screens, and downloading a document associated with the same session. In the second case, distributing addresses between steps can disrupt the operation. The design should separate independent jobs from processes that require continuity, using different rules when necessary. This distinction makes it possible to adopt random selection where it actually helps while preserving predictable behavior for activities with state, authentication, or dependencies between successive requests.

INGRESS AND EGRESSIPv4:PortPublic outbound IPv6
ApplicationConfigures IP:Port
Public IPv4 ingress198.51.100.10:8080
Public IPv6 egress2001:db8::24
The application connects to the same public IPv4 and ingress port. The proxy establishes the outbound connection over IPv6: static retains the address, rotating changes it at the configured interval in minutes, and random changes it for every new request. The destination must support IPv6. Illustrative addresses and port.
02

What each new request means

The expression “rotation on every new request” must be interpreted according to the protocol and the contracted integration. An application can send multiple HTTP calls inside a persistent connection or an already established HTTPS tunnel. In those cases, the proxy does not necessarily choose another IPv6 address for every internal message. The actual selection boundary may be the creation of a connection or an operation recognized by the service. Expected behavior therefore needs to be validated with the library, browser, or agent used by the customer.

During acceptance testing, execute a small sequence of identified tasks and observe the outbound address at a controlled test destination. Record when a connection opens, which requests share it, and when it ends. Compare those events with the observed IPv6 address. The goal is not to force indiscriminate reconnections, which increase cost and latency, but to confirm whether distribution matches the actual work. If the application requires session affinity, consider static service or a timed policy instead of applying randomness to every step of the process.

03

IPv6 egress and IPv4 where needed

This product uses IPv6 egress to reach compatible destinations. The application retains the same public IPv4 and port to connect to the VLSM.io proxy. For each new request, the service selects a public outbound IPv6 different from the one used for the preceding request. This does not mean that the destination website accepts IPv4 inside IPv6 packets. There are two connections with distinct roles, and the service must explicitly support the combination selected for the project.

Map the destinations that still depend on IPv4 and maintain a suitable alternative for them. A portal may present a homepage reachable over IPv6 while forwarding authentication, files, or queries to other names without the same availability. Verify the complete path, including redirects and supporting services. The outbound address family should be selected using this map rather than assuming that an IPv6 record for the main domain guarantees compatibility across every subdomain or every feature offered by the system throughout the complete customer workflow.

TRAFFIC AND ADDRESSINGFixed IPv4 ingress. IPv6 per request.
Fixed IPv4 ingressOutbound IPv6
Request 1198.51.100.10:80802001:db8::24
Request 2198.51.100.10:80802001:db8::78
Request 3198.51.100.10:80802001:db8::109
The application retains 198.51.100.10:8080 as its IPv4 ingress. The destination sees the public outbound IPv6: fixed in static mode, changed at the minute-based interval in rotating mode, or different for each new request in random mode. The rotating example uses a 5-minute window. This illustration does not display live traffic.
04

Randomness, prefixes, and capacity

Random selection is not synonymous with a unique address on every execution. Depending on pool size, service policy, and task volume, an IPv6 address may appear again. It is also incorrect to treat every address within the same prefix as a completely independent source: a destination can evaluate address groups, sessions, accounts, and usage patterns together. Sizing should start from verifiable operational needs, with attention to continuity, the source pool made available, and the conditions that apply to each specific integration and destination.

The IPv6 address space allows large pools to be organized, but it does not remove physical or logical limits. Proxy processing, simultaneous connections, name resolution, destination capacity, and storage speed still influence the result. Expand the project in stages and monitor task completion time. If a queue grows while the destination responds more slowly, adding addresses may not solve the problem. The actual constraint must be identified before increasing concurrency, resources, or the available outbound pool simply because more addresses can be configured.

05

Selection compatibility with each tool

Build a simple matrix of the tools that will execute random tasks: protocol, name resolution, connection reuse, and observed selection unit. A collector that opens new connections may behave differently from a browser maintaining persistent connections for multiple resources. With HTTPS forwarded through CONNECT, calls inside the tunnel should not be assumed to represent independent egress choices. Validating this matrix prevents adjustments to the IPv6 pool when the difference actually lies in a tool’s connection behavior or a reused configuration that was not designed for the new workload.

SOCKS5 clients may resolve names locally or use proxy-side resolution, identified as SOCKS5H in some tools. Choose the option compatible with the destination and validate the IPv6 result. UDP requires its own client and service capabilities, including association handling; accepting a SOCKS5 setting in the interface is not enough. When comparing addresses in tests, account for equivalent IPv6 representations that may use zero compression. Normalizing notation prevents two textual forms of the same address from being counted as different and distorting the distribution analysis.

06

Metrics that demonstrate results

Evaluate quality through the number of useful tasks completed, response integrity, and the time required to process them. The raw request count can increase even when an application repeats errors or receives pages without the expected data. Associate each task with a correlation identifier and record DNS, connection, TLS, and response failures separately. This organization makes it easier to identify whether a problem lies in the IPv6 path, the client, the session, or the content returned by the destination during a particular execution.

Use retry limits and progressively longer delays when transient failures occur. A task with an uncertain result should not be executed again without considering possible effects, especially when it creates records or changes data. For queries, use caching where appropriate and avoid repeating completed work. Random IP selection belongs to the connectivity layer; reliable automation also depends on queue management, duplicate identification, and clear decisions about when to wait, stop, or refer an exception for investigation by the people responsible for the workflow.

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

Growth with observable distribution

Monitor the pool at levels useful to the project: task, application, and provisioned address group. An aggregate view helps identify whether a few workers concentrate connections while others remain idle. However, do not interpret address variation as authorization to increase pressure on a destination. Concurrency limits should follow responses and access conditions. When throttling signals appear, slow the queue rather than turning random selection into an unlimited sequence of new attempts that consumes resources without producing useful results for the operation or its downstream users.

Sizing should consider connection renewal, TLS processing, response volume, and the ability to store results. A large IPv6 space does not remove those costs. Network infrastructure can be sized from 1 Mbps to multiple 100 Gbps according to the project; that range is not an individual promise per address or proxy.

Test growth with reproducible samples and record how latency and completed work change. If distribution improves without increasing task completion, investigate the next constraint before expanding again and adding resources that may not address the actual bottleneck.