Kosteneinsparungen durch Azure Reservierungen und Azure Sparpläne

Microsoft Azure is one of the leading solutions in the world of cloud computing platforms. Companies use Azure to host their applications, store data and provide scalable resources. It is important to optimize costs and maximize performance at the same time. In March, we already provided a comprehensive overview of where costs can be saved when licensing Microsoft software (including Azure: COMPLION | Profitability levers in Microsoft licensing - Part 2 (Azure). In this blog post, we would like to take a closer look at Azure reservations and Azure savings plans.

Azure reservations can be a way to save money by opting for one-year or three-year plans. This commitment gives the customer a discount on the resources they use. When the customer opts for a reservation, they commit to using that specific type of compute instance or instance family in a specific Azure region for a set period of time. For example, the customer could commit to using a "D4 v2 VM" in the Germany North region for one year. Reservations are well suited for workloads that run continuously and are stable. If the resources are fully utilized, reservations offer an opportunity to save costs.

According to Microsoft's original planning, the availability of the exchange of Azure compute reservations for Azure Virtual Machine, Azure Dedicated Host and Azure App Service was to end on January 1, 2024. When a compute reservation is exchanged, existing reservations are converted into new compute reservations. However, this availability of the exchange has now been extended by Microsoft until further notice. There is therefore no deadline for the exchange of Azure compute reservations at the time of publication of this blog post. Customers can therefore continue to exchange their compute reservations for different instances and regions until Microsoft notifies its customers again, which should be at least six months in advance.

The policy for canceling a reserved instance does not change. However, the total value of the canceled reservation may not exceed USD 50,000 for a billing profile or a single registration within a rolling 12-month period.

When a reservation is traded in, existing reservations are converted into a new savings plan. Customers can always exchange their reserved Azure instances for Compute for a savings plan. There is no time limit for a trade-in for a savings plan.

Azure savings plans are a similar - but slightly different - way of saving costs with Azure. With a savings plan, the customer commits to a fixed hourly cost for eligible compute services in all Azure regions for a specified period of time. Savings plans are therefore suitable for dynamic workloads that evolve. If the applications use different instance families or compute services or run in different data centers, savings plans are a better choice than reservations. The benefits of the savings plan apply to the entire use of participating compute services worldwide and are not limited to a specific region.

The decision as to whether a reservation or a savings plan is the better option depends on the customer's specific requirements. The following can be roughly summarized:

Overall, both reservations and savings plans offer a way to optimize costs in Azure. The choice depends on the customer's individual situation. In general, it is important to mention at this point that cloud cost optimization is a continuous process that requires regular reviews and adjustments.

Customers should also check how often they have exchanged Azure Compute reservations to date. In the case of regular exchanges, it could make sense to review the processes and adjust them if necessary. Finally, it is important to keep an eye on the policy for exchanging Azure reservations in order to exchange the reservations one last time when an exchange deadline is announced.

Author: Lennart Hollweg