One of our customers has undertaken a large on-premise to Azure migration over the course of the last 5 years. They are now at a point of relative stabilization and now that they have completed the migration and they have enabled users to be successful with more agile infrastructure and applications the customers focus has shifted to the sustainability of costs.
Cost has not been a core focus during the migration but as often happens with cloud migrations the costs are higher than the customer had originally planned for. One of the core reasons for the higher than expected costs is that a lot of the resources were over provisioned during their setup.
It is very easy to setup a virtual machine to be configured based on an estimated usage and then the team moves onto the next steps in the migration and no one has time to look back to applications that have been migrated and to check if the infrastructure is being well utilized.
In Azure one of the easiest ways to reduce your costs is to take advantage of Reservations where you can commit to using an amount of a resource type for a given period and then you can get a discount from Microsoft. Reservations are available in the Azure portal but they are not easily accessible and are disconnected from the resource owners. This is the main reason so many customers miss reservation opportunities.
When the customer begun using Turbo360 one of the initial tasks is modelling the costs in a way that makes sense for the organization. This customer chose to model cost based on the operations teams who are running the applications. This allows the operations teams to see their own costs and also any reservations that relate to their application as shown below.
The operations team can then choose the reservations that make sense for them and can plan to get the reservation applied in Azure which will show up the reservation as being applied in Turbo360.
The customer could get each of their teams to plan getting their appropriate reservations applied and without needing to make any changes to applications this reduces the customers costs significantly.
In this case it works out at $700k per year of cost savings for this customer.
Over the course of the next year the customer now has an additional $700k they can use to fund some projects to help with rightsizing resources, making some architecture changes and tuning activities which will further reduce the costs.
Easy wins like reservations are a good way to support reallocating costs for further savings.