# Premium Service Bus Namespaces for Low-Throughput or Non-Production Workloads Canonical: https://www.pointfive.co/efficiency-hub/inefficiencies/premium-service-bus-namespaces-for-low-throughput-or-non-production-workloads Azure Service Bus Premium namespaces run on dedicated messaging units (MUs) and are billed a fixed hourly rate per MU, regardless of how many messages... By: PointFive Updated: 2026-09-28 [Cloud Efficiency Hub](https://www.pointfive.co/efficiency-hub) The short version Azure Service Bus Premium namespaces run on dedicated messaging units (MUs) and are billed a fixed hourly rate per MU, regardless of how many messages flow through them. PointFive Research Cloud cost research at PointFive Azure service [Azure Service Bus](https://www.pointfive.co/efficiency-hub/cloud-services/azure-service-bus) Category [Other](https://www.pointfive.co/efficiency-hub/service-category/other) Reference CER-0454 Type Suboptimal Tier or SKU ## Explanation Why the waste happens and who it affects. A namespace that handles a few thousand messages a day costs the same per MU as one running near capacity. Standard namespaces, by contrast, run on shared capacity and are billed per operation on top of a base charge, so low-volume queues and topics usually cost less there. Premium is often chosen by default in templates or copied from production into development, test and staging subscriptions, where the Premium-only capabilities (dedicated capacity, messages up to 100 MB, virtual network integration and private endpoints, Geo-Replication, JMS 2.0, customer-managed keys) are not needed. A related overspend is Premium namespaces left at 2, 4 or more MUs after load has dropped. Microsoft's Well-Architected guidance for Service Bus calls out both: choose the right tier (Standard offers cost savings for non-production environments) and rightsize allocated resources. ## Billing model The pricing dimensions that drive this cost. Service Bus tiers use different pricing models. Premium messaging units Fixed hourly charge per MU; a namespace can have 1, 2, 4, 8 or 16 MUs, and CPU and memory use is not priced separately Hourly peak billing Premium is billed on the highest number of MUs allocated during each hour, so scaling down within the hour does not reduce that hour's charge Standard base charge and operations Standard namespaces pay an hourly base charge that includes a monthly allotment of operations, then a per-million-operations rate beyond it Brokered connections Included in the Standard base charge up to a documented limit, and not charged on Premium ## How to detect 5 checks to find it in your estate. - List namespaces with sku.name Premium (Azure Resource Graph or az servicebus namespace list) and flag those in subscriptions, resource groups or tags that identify development, test or staging environments - For each Premium namespace, review the CPU (NamespaceCpuUsage) and Memory Usage (NamespaceMemoryUsage) metrics over several weeks; Microsoft's sizing guidance says that if CPU usage is below 25 percent you might be able to scale down the number of MUs - Review Incoming Messages, Outgoing Messages and Incoming Requests to estimate the monthly operation count the namespace would generate on Standard pricing - Check whether the namespace actually uses Premium-only features: private endpoints or virtual network rules, Geo-Replication, queues or topics with a max message size above 256 KB, JMS 2.0 clients, or customer-managed keys - On the namespace Scale page, check whether it is on Manual scale with more than one MU and no autoscale rules ## How to fix 6 ways to remove the waste. - Move non-production and low-volume namespaces that do not need Premium features to Standard; Microsoft only provides a migration path from Standard to Premium, so downgrading means creating a new Standard namespace, recreating queues, topics, subscriptions and rules, draining messages and repointing clients to the new connection string - Before switching, compare the Standard cost at the observed operation volume with one Premium MU; at sustained high volume, per-operation charges can exceed the fixed MU price - Confirm that the workload can accept Standard limitations: no private endpoints or network service endpoints, 256 KB message size, variable latency on shared capacity and only the JMS 1.1 subset - For namespaces that must stay on Premium, lower the MU count or configure Custom autoscale on the Scale page with scale-in rules on CPU and memory; Microsoft recommends a scale-down cool-down of more than 30 minutes because billing is based on the hourly peak - Consolidate underutilized Premium namespaces where security and operational boundaries allow, so fewer MUs serve more entities - Add an Azure Policy or template default that restricts the Premium SKU to production subscriptions ## Documentation Vendor references for pricing and configuration. - [Azure Service Bus Premium Messaging Features learn.microsoft.com](https://learn.microsoft.com/en-us/azure/service-bus-messaging/service-bus-premium-messaging) - [Pricing - Service Bus azure.microsoft.com](https://azure.microsoft.com/en-us/pricing/details/service-bus/) - [Architecture Best Practices for Azure Service Bus learn.microsoft.com](https://learn.microsoft.com/en-us/azure/well-architected/service-guides/azure-service-bus) - [Azure Service Bus - Automatically update messaging units learn.microsoft.com](https://learn.microsoft.com/en-us/azure/service-bus-messaging/automate-update-messaging-units) - [Monitoring data reference for Azure Service Bus learn.microsoft.com](https://learn.microsoft.com/en-us/azure/service-bus-messaging/monitor-service-bus-reference) - [Migrate Azure Service Bus namespaces - standard to premium learn.microsoft.com](https://learn.microsoft.com/en-us/azure/service-bus-messaging/service-bus-migrate-standard-premium) ## Related inefficiencies [Browse the library](https://www.pointfive.co/efficiency-hub) - Microsoft Fabric CER-0337 ### [Oversized Microsoft Fabric Capacity During Low-Utilization Periods](https://www.pointfive.co/efficiency-hub/inefficiencies/oversized-microsoft-fabric-capacity-during-low-utilization-periods) Microsoft Fabric capacity is billed based on the provisioned SKU tier (F2, F4, F8, F16, F64, etc.) on an hourly basis, regardless of whether workloads are actively consuming those resources. Each SKU tier provides a fixed pool of Capacity... Other - Azure Container Registry CER-0329 ### [Overprovisioned or Idle Azure Container Registry Tier](https://www.pointfive.co/efficiency-hub/inefficiencies/overprovisioned-or-idle-azure-container-registry-tier) Azure Container Registry charges a fixed daily fee based on the selected tier - Basic, Standard, or Premium - regardless of whether the registry is actively used. This means a registry with zero image pulls, zero pushes, and no active... Other - Azure Logic Apps CER-0309 ### [Idle or Untriggered Azure Logic Apps Generating Continuous Charges](https://www.pointfive.co/efficiency-hub/inefficiencies/idle-or-untriggered-azure-logic-apps-generating-continuous-charges) Azure Logic Apps can quietly accumulate costs even when no workflows are actively executing, but the mechanism differs significantly depending on the deployment model. In the Consumption (multitenant) plan, Logic Apps with polling triggers... Other --- Source: the public page above. Product screenshots and illustrative interfaces are examples, not live customer data.