There has been a lot of talk surrounding Microsoft’s Azure Reserved VM Instances (RIs) lately, and with good reason.
The opportunity to purchase Azure compute at lower prices than current agreements has caused many organisations to get more than a little excited. AWS has has long had EC2 reserved instances available, so the Microsoft release has attracted a lot of attention.
However, before you jump on board and enable this new feature, you need a clear picture of how RIs will affect your organisation.
Here we’ll explain how RIs work, and discuss the benefits and drawbacks to help you decide if enabling them is the right move for your organisation.
When you sign up for RIs, you are reserving and purchasing virtual machines (VMs) in advance. The one or three-year terms on offer require up-front payment, and a commitment to a designated number of VMs for the duration of your term. While you can exchange reservations without a fee, cancellations will incur a fee (12%) and there are limits on how much can be cancelled each year.
For increased flexibility, you can combine RIs with pay-as-you-go prices. This helps you manage costs more effectively across predictable and variable workloads.
The key benefit being pitched by Microsoft is the chance to significantly reduce your cloud computing costs, by up to 72 percent. While there is a great opportunity to save for many organisations, there are others whose circumstances require a different approach. For example, some companies would benefit more from using higher performing servers in Azure at lower costs than they are currently paying now.
To take advantage of the potential cost savings of RIs, your organisation needs to meet certain criteria. To be sure you will benefit, it’s important that you carefully assess your current situation and factor in the key considerations before you make the call to switch.
Your RIs are fully pre-paid, so you will need to include the payment of the full one or three-year commitment in your budget for the first month. Plus, if your business or system requirements change, as mentioned above there is a cancellation fee of up to 12 percent which can be a substantial amount. RIs are designed to deliver maximum savings for any VM that’s running 24×7 x365, and that are unlikely to change in size in the future.
Currently, RIs are not assigned to specific VMs, which makes mapping your charge back to business units difficult. Be sure to work closely with any departments that may be affected by the implementation, so that your decision considers the entire company.
Tips for Success
To manage RIs effectively, it’s important to plan your transition carefully to deliver maximum value. Consult the business owners of systems to ensure long-term run time of projects, and confirm with budget owners and finance that pre-payment of RIs will fit in with your current models. In the cloud based consumption world, having a full 12-month or 3-year commitment paid up front may not fit the budget requirements of the business.
You will also need to update your reporting capabilities, so you can track the usage and provide visibility to the business. Plus, you will need to continue to monitor and plan for RIs on an ongoing basis to ensure they remain cost-effective moving forward.
Are RIs the Right Move for Your Organisation?
At the end of the day, the key question is: Will RIs save my organisation money? To find out, you will need to assess the following:
- your current cloud expenses
- workloads that can be moved to use RIs
- the up-front cost to turn on RIs
- year one through to year three cost comparisons.
Completing this step accurately is crucial to discovering if enabling RIs across your workloads, in combination with pay-as-you-go prices, or taking a different approach, is the best move for your organisation.
Our initial customer assessments of RIs indicate a potential saving of ~$100,000 per annum per 100 VMs.