An Introduction to Infrahub Resource Manager (With FAQs)

Q: What happens when a pool runs out of resources?

When the Resource Manager attempts to allocate from an exhausted pool, the allocation request fails with an error indicating that no resources are available. Your automation should handle this error appropriately, either by alerting operators, failing safely, or trying an alternative pool. You'll need to expand the pool by adding additional source resources before successful allocations can resume.

Wim Van Deun

Jan 19, 2026

Infrahub Resource Manager in a nutshell

Infrahub Resource Manager automates the allocation of network resources from predefined pools, eliminating manual assignment and preventing conflicts across your infrastructure.

Instead of manually tracking which IP addresses, VLAN IDs, or BGP ASNs are available, you define pools of resources and let the Resource Manager handle allocation automatically.

When you need a loopback IP for a device or a /31 subnet for a point-to-point link, the Resource Manager instantly provides the next available resource, tracks the allocation, and ensures no other system can assign that same resource.

This eliminates tedious lookup work, prevents duplicate assignment mistakes, and provides complete visibility into what's allocated and what's available.

Infrahub Resource Manager is designed for automation from the ground up, making it perfect for modern infrastructure-as-code workflows. And because it's branch-agnostic, you can safely allocate resources in development branches knowing those allocations are immediately reserved across your entire system, preventing merge conflicts before they happen.

Resources you can manage with Infrahub Resource Manager

Infrahub Resource Manager handles three primary types of resource pools, each designed for specific infrastructure needs.

IP address pools allocate individual IP addresses from defined IP prefixes. Perfect for device loopback interfaces, management IPs, or service endpoints.
IP prefix pools allocate smaller subnets from larger parent prefixes. This is useful for carving up address space into smaller chunks, like allocating /31 point-to-point subnets from a /24 block, or providing customer allocations from your provider-assigned space.
Number pools allocate sequential numeric identifiers from defined ranges. While network-specific uses like VLAN IDs or BGP AS numbers are obvious, Number Pools are remarkably flexible. You can use them to manage virtually any numeric identifier, such as circuit IDs, interface indices, or service contract numbers.

IPAM in Infrahub

The IP address management module in Infrahub Resource Manager has all the capabilities you would expect for IPAM.

At the foundation are namespaces allowing you to maintain separate, potentially overlapping address spaces. For example, you might have "Production" and "Development" namespaces, each containing 10.0.0.0/8 space without conflict.

Within namespaces, Infrahub manages IP prefixes in a hierarchical structure. The system automatically understands relationships between parent and child prefixes—that 10.1.1.0/24 is contained within 10.1.0.0/16, for instance.

You can define the member type for any prefix (prefix or address), and Infrahub uses this to compute utilization appropriately, tracking allocated addresses for address-type prefixes and child prefix allocations for prefix-type prefixes. This provides accurate visibility into capacity at every level of your addressing hierarchy.

How is Infrahub Resource Manager different from IPAM tools?

Traditional IPAM tools focus primarily on tracking IP address space. They’re essentially sophisticated databases for recording which IPs are allocated.

Infrahub Resource Manager can manage IP addresses and IP prefixes but many other resources as well, including VLAN IDs and BGP ASNs. In fact, the Resource Manager can manage virtually any sequentially numbered resource you want to track. The same allocation mechanisms, visibility tools, and automation patterns work regardless of resource type.

The Resource Manager is designed for automation and can be called by workflows, Infrahub Generators, and CI/CD pipelines. Its idempotent allocation behavior means automation can safely request resources without creating duplicates when code runs multiple times.

The integration with Infrahub's branching model is another key differentiator. Resource allocations are branch-agnostic. A resource allocated in any branch is immediately reserved across all branches, preventing duplicate allocations that would cause merge conflicts. Traditional IPAM tools have no concept of this workflow.

Common use cases for Infrahub Resource Manager

Device provisioning workflows can automatically assign loopback IPs, management IPs, and router IDs from appropriate pools during device creation, eliminating manual lookup and ensuring consistency.
Data center expansion and point-to-point link creation benefit from automatic prefix allocation. As you add switches or routers needing /31 links, the Resource Manager allocates the next available subnet, ensuring no overlaps.
Customer or tenant onboarding in service provider environments is streamlined by automatically allocating IP prefixes and BGP AS numbers during onboarding workflows.
VLAN management across campus or data center networks ensures unique VLAN allocation and prevents conflicts, especially valuable when different teams manage different network parts.

How Infrahub Resource Manager works

The Resource Manager operates on a straightforward principle: You create pools that reference source resources, then request allocations either directly or during infrastructure object creation.

For IP address and IP prefix Pools, you attach one or more prefixes as the source resources from which allocations will be drawn. When creating an IP prefix pool, you also set a Default Prefix Length, which determines the size of subnets that will be allocated from the pool (like /31 for point-to-point links or /24 for site networks). This default can be overridden on a case by case basis if you need a different size for specific allocations.

For a number pool, you define the available range (like 100-200), then the pool is bound to a node’s attribute (e.g., VLAN > VLAN ID).

Allocation happens through two methods:

Direct allocation creates standalone resource objects from a pool, which is useful for pre-provisioning resources or building inventories.
Relationship resource allocation happens during infrastructure object creation. When creating a router, for example, you request its loopback IP through a relationship to a pool, and the Resource Manager allocates and establishes the relationship atomically.

You can request allocations through the Infrahub GUI, the Python SDK with methods like allocate_next_ip_address(), or GraphQL with mutations like IPAddressPoolGetResource. The Resource Manager identifies the next available resource, marks it as allocated, and records metadata including which branch allocated it and any identifier you provided.

How Resource Manager works with Infrahub's schema system

The Resource Manager is built on Infrahub’s schema system, which means it inherits the full flexibility provided by schemas. The core pool types—CoreIPAddressPool, CoreIPPrefixPool, and CoreNumberPool—are defined in Infrahub’s schema just like any other object type.

Because pools are first-class schema objects, you can create relationships from any node to a pool. For example, you can link a Site node to an IP prefix pool to explicitly document which pool should be consumed for that site. This creates an explicit association that eliminates the need to rely on parsing names or descriptions.

These relationships can be used by your automation logic to determine which pool to draw from when provisioning new resources for a given site.

Understanding idempotent allocation

Idempotency means performing an operation multiple times produces the same result as performing it once. In the Resource Manager, idempotent allocation means requesting a resource with the same identifier multiple times returns the same resource rather than creating duplicates.

Idempotent allocation is critical for automation. Consider a script that provisions a router and allocates a loopback IP. If the script runs but your deployment system retries it due to a timeout, you don't want a second IP—you want the same one already allocated.

The Resource Manager provides idempotency through the identifier parameter. When you request an allocation, you can include an identifier, a string that uniquely represents the allocation's purpose. For example, when allocating a loopback IP for router edge-01, you could use identifier edge-01-loopback. If the Resource Manager finds an existing allocation with that identifier, it returns it. Otherwise, it allocates a new resource and records the identifier.

In this way, your automation can be truly declarative, dramatically simplifying automation and testing.

Using weighted allocation

Weighted allocation gives you control over the order in which the Resource Manager draws resources from a pool. Each source can have an allocation_weight attribute. The Resource Manager allocates from higher-weighted sources first, exhausting them before moving to lower-weighted ones.

Let’s say you have an IP address pool drawing from 10.100.0.0/24 and want to transition to 192.168.200.0/24. Rather than abruptly cutting over, you add the new prefix to your pool but give the old prefix a higher allocation weight (say, 10) and the new prefix a lower weight (say, 5).

The Resource Manager continues allocating from the old prefix until it's exhausted, then automatically starts allocating from the new prefix. This provides a gradual, automatic migration without needing to change any of your allocation logic or create and swap between different pools.

Weighted allocation is also valuable for capacity planning and staged rollouts. If you're anticipating growth and want to preserve address space in a particular block for future use, you can add that block to your pool with a lower weight. The Resource Manager won't use it until your higher-weighted blocks are exhausted, effectively reserving that space while still making it available as a safety net if needed.

Infrahub Resource Manager is branch-agnostic

One of the Resource Manager's most powerful features is its branch-agnostic allocation, which solves resource allocation conflicts during merges.

As an example, let’s say two engineers are working in separate branches, and both are allocating loopback IPs from the same pool. In typical systems, each branch allocates the resources independently, potentially assigning the same IP to different routers. When branches merge, you have duplicate IPs, which are often discovered only after deployment.

The Resource Manager prevents this entirely. When a resource is allocated in any branch, that allocation is immediately visible and reserved across all branches. If one engineer allocates 192.0.2.1, the other cannot—they get 192.0.2.2 instead. When branches merge, there are no conflicts because resources were never duplicated.

Integrating Infrahub Resource Manager into your automation workflows

Infrahub Resource Manager is designed to be a first-class component of infrastructure automation, with multiple integration points that fit different automation scenarios.

The Resource Manager can be accessed through the Python SDK and GraphQL API, making it the foundation for various automation integrations:

The Python SDK provides intuitive methods like allocate_next_ip_address() that handle GraphQL complexity behind the scenes, making it easy to integrate allocation into Python-based automation tools.
The GraphQL API offers maximum flexibility for custom automation or non-Python tools. Mutations like IPAddressPoolGetResource provide direct access to allocation functionality from any language or tool.

Using these two entry points, you can integrate the Resource Manager into:

Infrahub Generators that automatically allocate resources during execution and incorporate them into generated configurations.
CI/CD pipelines that allocate resources as part of infrastructure deployment workflows—for example, allocating IP addresses or VLAN IDs before provisioning new services.
Ansible playbooks, Terraform modules, or custom scripts that need to request and track infrastructure resources programmatically.

Infrahub Resource Manager FAQs

How can I see what resources are in use and what's available in Infrahub?

You can see resource utilization at the pool level through the Infrahub web interface, which shows all allocated resources, utilization percentages, and available capacity. Each allocation includes details about what was allocated, when, which branch allocated it, and any identifier used. You can also query this information programmatically via GraphQL or the Python SDK.

What happens when a pool runs out of resources?

When the Resource Manager tries to allocate from an exhausted pool, the allocation request fails with an error indicating no resources are available. Your automation should handle this error appropriately, either by alerting operators, failing safely, or trying an alternative pool. You'll need to expand the pool by adding additional source resources before successful allocations can resume.

Can I manually override or reserve specific resources within a pool?

Yes, you can manually allocate specific resources from a pool rather than taking the next available one. This is useful for reserving particular IPs or numbers for specific purposes. Once manually allocated, that resource is marked as used and won't be automatically allocated to other requests.

Can I release or reclaim allocated resources?

Yes, allocated resources can be deleted to return them to the available pool. You should ensure the resource is no longer in use in any configurations or systems before releasing it to avoid creating issues. The Resource Manager doesn't automatically track whether an allocated resource is actively deployed.

What happens to allocations when I delete or merge a branch?

Allocations remain in place when branches are deleted or merged. When you merge a branch, any allocations made in that branch become part of the target branch. You must explicitly release allocations if you want to reclaim those resources.

Can I exclude certain addresses from automatic allocation?

Yes, you can exclude specific addresses by manually allocating them before automatic allocation begins, or by structuring your pool sources to not include those addresses.

Can multiple pools allocate from the same parent resource?

Yes, multiple pools can reference the same parent prefix or number range as their source. This is useful when you need different allocation behaviors or want to segment allocations for different purposes while drawing from the same address space.

How do I handle multi-site or multi-region resource allocation?

Your resource pool structure should mirror your IP addressing strategy. If you use global address space (like a single management prefix across all sites), create one resource pool that allocates from that global space. If you use site-specific address space (like separate production prefixes per site), create separate resource pools for each site. This ensures allocations come from the appropriate address space for each location.

Can I use Infrahub Resource Manager with custom object types?

Yes, the Resource Manager is schema-based and extensible. You can create custom object types (like a Site or ServiceContract node) that have relationships to resource pools, enabling automatic resource allocation when those objects are created. This allows you to integrate the Resource Manager into your specific data model and workflows.

Can I see allocation history or audit trails for compliance?

The Resource Manager tracks metadata about each allocation including when it was allocated and which branch allocated it. For more comprehensive audit trails, you can use Infrahub's broader audit and versioning capabilities, which track all changes to objects including resource allocations, providing a complete history for compliance purposes.

How do I migrate existing resources into the Resource Manager?

You can import existing manually-assigned resources by creating the corresponding resource objects in Infrahub and associating them with your pools. For IP addresses, for example, you would create IP address objects linked to your prefixes. Once imported, these resources are marked as allocated and the Resource Manager won't reassign them.

What are the limitations or constraints of Infrahub Resource Manager I should know about?

The Resource Manager allocates the next available resource deterministically but doesn't currently support complex allocation strategies like prefix fragmentation. There's no automatic scaling so pools must be explicitly expanded when exhausted. The Resource Manager also doesn't track whether allocated resources are actively deployed versus simply reserved, so you’ll need to clean up any unused allocations.

Ready to try Infrahub Resource Manager?

Follow a hands-on tutorial: Our technical deep dive on the Resource Manager walks you through implementing automated resource allocation step-by-step, with real code examples and complete workflows.
Implement the Resource Manager in your environment: The Resource Manager implementation guide in our documentation provides everything you need to set up pools, configure allocation, and integrate with your automation.
See it in action: Play in the sandbox or request a personalized demo to see how the Resource Manager can eliminate resource conflicts and streamline allocation in your specific infrastructure environment.