5 FinOps Techniques for Cloud Cost Optimisation using VMWare CloudHealth
Managing your Cloud expenses can sometimes feel like a bit of a black art. Especially now, as we see the growing FinOps community sharing and learning about cloud cost optimisation techniques, that magical art of analysing, taming and managing your Cloud spend is evolving as quickly as the Cloud technologies that it supports. At nubeGo, we like to keep things clear and understandable, so we encourage our Razor and NCMS Cloud Managed Service customers to make use of VMWare Cloudhealth to monitor and optimise their Cloud costs, security and governance.
nubeGo team member Emma Button recently achieved her CloudHealth Platform Administrator certification. In this blog post, Emma shares her top 5 techniques for Cloud Cost Optimisation in CloudHealth.
1. Define and enforce a tagging policy for Cloud assets In order to gain the greatest understanding of your cloud estate, and to help you accurately allocate costs to different groups of people, it is important to tag your Cloud resources. OK, not all Cloud resources are taggable, but if you can… you should. At a minimum, make sure you add an “owner” tag when you create a new Cloud resource - this will help you trace who the asset belongs to. Other tags you might wish to add include “department”, “application”, “cost center” or maybe even “version” - anything that allows you to group and categorise our cost data so that you can better understand where costs are coming from. Using CloudHealth, you can build Perspectives from tags and accounts that allow you to conduct charting and analysis for a specific group or groups of assets. Try using a tool such as AWS Config (or CloudHealth governance policies) to automatically alert on untagged resources, or even to enforce compliance by preventing untagged resources from being created.
2. Make use of Reserved Instances where possible If you are making use of Cloud compute instances such as Amazon EC2, consider whether a reservation of compute credits would benefit you. If you’re running multiple EC2 workloads, or long-running instances of a similar type then in all likelihood, a fixed-term reservation will save you money. The art of reservation planning, forecasting and management merits a blog post of its own, but to make things easy, CloudHealth provides a number of wizard and chart tools which automate the process of assessing your reservation needs.
Use the EC2 RI Analyzer to understand how your RIs are currently being used and to find out which new reservation purchases CloudHealth recommends, based on your usage
Use the EC2 RI Optimizer to generate quotes for new Reservations and understand the potential cost savings over time, including amortization
3. Monitor and alert on budgets Most Cloud providers give you the ability to alert when an overarching budget is approaching a threshold. FinOps tools such as CloudHealth give you the ability to perform far more detailed budgeting and apply more granular allocation of budget allowances, and alerts. Within CloudHealth you can define a budget for your overall Cloud spend, or for a group of assets or accounts (using the Perspectives I mentioned earlier). You can visualise how your groups are tracking towards their budgets and you can define rule-based alerts to notify the right people when limits are being approached.
4. Plan for Data Transfer Costs One of the “hidden” costs that we find people often forget to include in their existing cloud budgets is the cost of transfer for data out of the Cloud, or between Cloud regions. CloudHealth provides us with a Data Transfer usage report which makes it easy to see and understand the data transfer costs we incur, and by which service. Armed with this information, we can make informed decisions about application architectures, network topologies and storage choices.
5. Delete unused storage volumes Perhaps the biggest quick-win for any one-off or ongoing Cloud optimisation initiative is to remove unused resources. Unused storage volumes are a common type of Cloud asset that can be easily overlooked. Using CloudHealth to create a monitoring policy, you can alert when a storage volume has been unattached for a certain amount of time. And, with the correct configuration, you can even create a policy which will automatically delete unattached volumes, given your approval. We do a similar thing use automatic AWS Config rules in our own NCMS managed platform for AWS so we know that small changes can result in big cost savings.
As VMWare partners, nubeGo are pleased to be able to offer VMWare Cloudhealth on a monthly subscription as part of our Razor cost optimisation programme - it also comes as standard for our Cloud Managed Service clients. Get in touch with a member of the team if you think it could help you monitor and optimise your cloud spend.