Talk to a specialist

BLOG VLSM.io

Leasing IPv6 subnets: plan your expansion

Plan IPv6 subnets around environments, responsibilities, and growth, using the address space without confusing it with traffic capacity.

IPv6-only egressConnectivity and intelligence
Globe with IPv6-only outbound connections
IPv6-only egress2001:db8::24Illustrative addresses
Estimated reading time: 8 min8 sections
Leasing IPv6 subnets: plan your expansion
Leasing IPv6 subnets: plan your expansion
01

Plan the hierarchy before counting addresses

IPv6 address space allows broader network organization, but abundance does not remove the need for planning. Instead of starting with individual address count, describe regions, sites, environments, customers, or services requiring separation. Define which divisions help routing, administration, and incident investigation.

VLSM.io’s IPv6 subnet rental considers prefixes from /56 through /29, subject to project requirements and availability. That range covers very different scales; the choice should reflect a usage and expansion structure your team can administer and explain, rather than the largest number available.

A useful hierarchy can reserve portions for environments and retain nearby room for expansion, avoiding constant reassignment. That does not mean copying a universal model: a company with a few locations and a platform serving many customers have different needs. Document the meaning of each level and avoid divisions understood by only one person. When a new service appears, the team should know where it belongs and who approves its assignment. That predictability supports inventory automation and reduces conflicts between projects expanding at the same time.

CONNECTIONS IN MOTIONThe path of your connection
ApplicationJob or integration
ConnectivityIPv6-only egress
DestinationSystem or service
The illustrated service uses IPv6-only Internet egress. IPv4 egress requires a separate service. The animation does not represent live traffic.
02

Count subnets, not only addresses

An IPv6 address has 128 bits. To determine how many /q subnets fit inside a /p block, where q is at least p, calculate two raised to q minus p. A /56 contains 256 /64 subnets; a /48 contains 65,536; a /29 contains 34,359,738,368.

These values describe mathematical subdivisions, not a promise that infrastructure can activate or administer all of them simultaneously. /64 is common for IPv6 local networks, while the complete plan must consider requirements for each link type and configuration mechanism.

Do not automatically carry over the conventional IPv4 host calculation that subtracts network and broadcast addresses. IPv6 does not use broadcast in the same way, and address-space organization follows different criteria. Analyze prefix size together with addressing, routing, and applications.

For example, a team may need several clearly identified segments despite using few devices in each. Reducing the plan to a machine count loses the hierarchy’s value. Conversely, an enormous block does not fix software unable to represent or store an IPv6 address correctly in its configuration and records.

PrefixContained /64 subnets
/56256
/4865,536
/2934,359,738,368
03

Prepare applications and IPv4 coexistence

Receiving an IPv6 block does not make every destination compatible with the new family. Inventory authentication, APIs, files, monitoring, and support dependencies, checking each required path. The contracted IPv6 subnet provides IPv6-only egress. If the operation also needs IPv4 egress, contract another service and monitor each path independently.

A failure can remain hidden when a client automatically finds another connection option. To validate deployment, observe the family actually used and test components under controlled conditions. Coexistence should be an explicit decision with clear routes and responsibilities, rather than a condition discovered during an outage affecting only part of the application.

Review configuration fields, address storage, form validation, and operational reports. A field sized only for IPv4 notation may truncate data or reject a valid entry. IPv6 addresses can have equivalent textual representations, so comparison and deduplication require consistent handling. Use suitable libraries to interpret addresses instead of improvised expressions. Also test third-party systems receiving your records. Expansion becomes easier when inventory and support tools understand IPv6 from the beginning, without manual conversions in every support request or separate spreadsheets used to compensate for software limitations.

TRAFFIC AND ADDRESSINGOne IPv6 service. IPv6-only egress.
Source2001:db8::24
Destination2001:db8::20
Return2001:db8::24
The outbound path of this service uses IPv6 only. Reaching an IPv4 destination requires a separate IPv4 service; there is no dual-stack egress. Illustrative addresses.
04

Match delivery to the control your team needs

Delivery mode defines how the block participates in infrastructure. A proxy provides an application interface; GRE and VXLAN support network designs; cross-connect establishes a suitable physical interconnection; VM and bare metal allow components to run in a contracted environment. These choices require different levels of configuration and responsibility.

Describe which systems are inside or outside the datacenters and which paths are needed. Also confirm the family used to access the service and to reach the destination, because IPv4 ingress can coexist with IPv6 egress while remaining separate legs.

For routed connections, validate forward and return paths, filters, and supported packet sizing. ICMPv6 participates in functions necessary for network operation; indiscriminate blocking can break important mechanisms, including packet-size information. Policy should be specific, preserving necessary messages while controlling what should cross each boundary. Do not copy IPv4 rules without reviewing their suitability. Include transfers of different sizes in testing, because a successful initial connection does not demonstrate that all application traffic will work over the configured path, especially when encapsulation is involved.

When using a proxy, confirm supported protocols and how the application specifies its destination. An HTTP configuration does not automatically cover UDP requirements, and a client’s name-resolution option can change where lookup occurs. Test the actual tool and its auxiliary components.

For tunnels, document endpoints and associated security policies. The block is only one part of the solution; delivery, routing, and client configuration must work together. This view prevents a large address allocation from being delivered without a usable path for the expected workload or a clear troubleshooting procedure.

05

Manage assignments and capacity separately

Maintain a source of truth for assigned prefixes, reservations, owners, and change dates. A customer or environment subdivision should have creation, review, and release rules. Automate overlap checks and record desired state before applying changes. Also establish how a migration updates DNS, access lists, and dependent configurations.

Available space allows planning with room to grow, but does not replace inventory discipline. Without it, teams lose the ability to locate resources and may retain old associations long after a project ends, making support and future reassignment harder.

Size transmission and processing capacity independently of the prefix. VLSM.io’s stated scalability, from 1 Mbps through multiple 100 Gbps, depends on design and contracting; it is not multiplied by the number of addresses inside a /56 or /29. Observe bandwidth, connections, packets, computing resources, and permitted destination workload.

A larger pool may organize operations but does not create capacity in an external service. For collection and RPA applications, maintain global work limits and monitor completed outcomes to identify whether the bottleneck is in the network or another stage.

06

Expand with evidence and clear ownership

Clarify who administers the resource, route announcements, and related records during use. A rental agreement does not automatically equal a registration transfer at an Internet registry. Where RPKI is involved, route-origin authorization should match the announcement design; it does not replace connectivity tests or application permissions.

Define contacts and procedures for maintenance, incidents, and termination. This makes expansion more predictable because new prefixes follow a known process instead of relying on improvised decisions during deployment or on knowledge held by one person.

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

Make the pilot represent the next stage

The pilot should demonstrate more than a responding address. Validate a complete assignment, its inventory record, routing, access policy, and an actual application using IPv6. Include failure recovery, a controlled change, and removal of a test assignment.

Compare the outcome with the expansion plan and adjust the hierarchy before multiplying environments. When presenting the project to VLSM.io, explain how many segments are needed, how they will be organized, and which capacity will be used. A well-planned deployment turns IPv6 space into manageable infrastructure ready to grow with clarity.

08

Institutional references

Reference documents on the topics covered in this article.