Azure Shared Capacity Reservations Guide

Key insights

• Capacity Reservation Groups (CRGs): CRGs hold reserved compute capacity defined by VM size, region, availability zone, and quantity.
  They are created on demand and a VM must specify the CRG to consume the reserved capacity.
• Capacity Guarantee vs Quota: A subscription quota only limits capacity requests, but does not promise availability.
  Capacity reservations provide an SLA-backed guarantee that matching VMs get the reserved compute when needed.
• Sharing Reservations: A provider subscription can share a CRG with consumer subscriptions by adding them to the CRG sharing profile.
  Only VMs in consumer subscriptions that match the reservation SKU, region, and zone will use the shared capacity.
• Centralized management and cost optimization: Sharing lets teams reuse capacity across subscriptions, reducing duplicated reservations and lowering costs.
  This model suits disaster recovery, dev/test, and multi-subscription governance while keeping capacity management centralized.
• Compatibility and migration: You can move running VMs into a reservation so they start using reserved capacity immediately.
  Reservations can work alongside Reserved Instances and Savings Plans, but review billing scope and rules for your subscriptions.
• Current limits and controls: The feature is in public preview and lacks a portal UI today, so use the CLI or APIs and plan for manual steps.
  Be aware of availability zone mapping issues, required permissions, and other preview limitations before you rely on shared reservations in production.

Introduction

The latest YouTube video from John Savill's [MVP] explains a new public preview capability in Azure called Shared Capacity Reservation Groups. In clear terms, the video walks viewers through what this feature does, why it matters, and how it changes capacity planning for multi-subscription environments. For editorial clarity, this article summarizes the video’s key points and frames the new feature in practical terms for IT decision makers and cloud engineers. Consequently, readers can quickly see how sharing reserved capacity could affect operations and costs.

What the Video Covers

First, the presenter reviews the basics of capacity reservations and contrasts them with subscription quotas, emphasizing that quota does not equal a capacity guarantee. Then, the video outlines the new sharing model that lets one subscription act as a central pool and grant access to other subscriptions, which the video describes as provider and consumer roles. In addition, the author walks viewers through practical topics like moving VMs into reservations, compatibility with reserved instance and savings plans, and current limitations. As a result, the audience gets both conceptual context and operational detail.

How Shared Capacity Reservations Work

At its core, a Capacity Reservation Group (CRG) defines a pool of reserved VM size capacity in a region or availability zone. The new sharing ability allows the owner subscription to explicitly add other subscriptions to the CRG so they can consume matching capacity when they deploy VMs. The video stresses that VMs must match SKU, region, and zone criteria and must be configured to reference the CRG to receive the reserved capacity. Therefore, correct mapping and deployment settings remain essential for the reservation to apply.

Moreover, the presenter highlights that these reservations provide an SLA-backed guarantee of capacity distinct from quota increases. While quotas raise the maximum allowed resources, reservations actually guarantee compute placement when the reservation exists. This guarantees availability for critical workloads, which is useful for disaster recovery and production scenarios. However, the guarantee only lasts while the reservation remains active and correctly associated.

New Scenarios and Benefits

Sharing reservations unlocks several new scenarios, including centralized capacity for service providers and cross-team disaster recovery where isolated subscriptions still access reserved capacity. Consequently, organizations can consolidate reservations to reduce idle capacity and improve utilization across environments such as development, testing, and DR. The presenter also points out that teams can keep separate subscriptions for compliance or billing while benefiting from a common compute pool. Thus, the feature supports governance and cost-efficiency at the same time.

Additionally, the video explains that reserved capacity can be repurposed dynamically, which helps maximize value compared with long-term reserved instance commitments. It also notes compatibility with savings plans and reserved instance approaches, so teams can combine cost models depending on predictability and scale. This flexibility supports both short-term operational needs and longer-term financial planning. Therefore, organizations must weigh immediate agility against longer-term committed discounts.

Tradeoffs and Operational Challenges

Despite the benefits, the presenter highlights several tradeoffs that organizations must consider. Centralizing capacity simplifies management, but it can also introduce single points of contention if permissions or AZ mapping are misconfigured, which may limit the usable capacity for some consumers. In other words, sharing reduces duplication but increases dependency on correct configuration and coordination. As a result, teams should balance centralized control with clear operational procedures and runbooks.

Another challenge is the current lack of a portal UI at the time of the preview, which means administrators often rely on CLI, ARM templates, or PowerShell to create and share reservations. This creates a higher bar for adoption until a graphical experience arrives. Also, the video calls out the importance of AZ mapping between subscriptions because zones must align for reservations to apply, which can complicate multi-region or multi-subscription designs. Therefore, planning and testing are essential to avoid surprises in production.

Finally, the video addresses compatibility concerns such as how moving running VMs into a reservation works and how cost models interact with reservations, reserved instances, and savings plans. Moving VMs may require actions that affect uptime depending on the method, so teams must weigh the operational disruption against the value of guaranteed capacity. Similarly, choosing between on-demand CRGs and long-term commitments involves trading flexibility for deeper discounts. In practice, teams should model scenarios and run pilot tests to find the right balance.

Permissions, Availability, and Next Steps

Permissions play a key role because the provider subscription must explicitly share CRGs with consumer subscriptions and grant appropriate access. The video emphasizes that access control and role assignments need careful design to prevent accidental consumption or misconfiguration. At the time of the preview shown, administrators may also find gaps in tooling and should prepare to use automation for repeatable deployments. As a result, organizations that move early should plan for additional governance and monitoring work.

Conclusion

In summary, the YouTube presentation by John Savill's [MVP] provides a clear, practical look at Shared Capacity Reservation Groups and the new possibilities they enable for Azure capacity management. The feature can reduce waste, centralize management, and ensure availability across subscriptions, but it also increases the need for careful mapping, permissions controls, and operational discipline. Therefore, teams should evaluate the tradeoffs between flexibility, cost, and control and run pilots before broad rollout. Ultimately, this preview points to a useful step forward in how organizations manage cloud capacity.

Keywords

Shared capacity reservations, Azure shared capacity reservations, capacity reservation best practices, reservation management, cloud cost optimization, reserved instances vs reservations, shared capacity pricing, capacity planning for reservations