5(+1) ways to control your cloud IT costs [part 2 -> Clean-up]

Cloud is gaining enormous momentum. Based on research that Devoteam did with IDC, its presence will only grow bigger in the coming years. Although cloud can be a cost saver, we see it the other way around on multiple occasions. In this post I want to give you tangible directions that you can start to apply directly.

Clean-up

During your work day you are working behind a desk for a huge part of the day. Having a clean desk saves you from distraction, makes you more efficient and provides focus. Besides that, it is secure, since there are no (confidential) documents available after you leave. In the cloud, it basically works the same and - as it is the case in your workplace - a clean-up can be done in various ways. When thinking about clean-up, organizations have to think about which resources are being used, how they are being used and when they are being used.

Within Clean-up we see the following ways to control your costs:

• Efficient use
• Shut down of unused resources
• Spot instances

Efficient use

Race cars and sports cars are very different as they were made for two different purposes. Sports cars are used for everyday driving and are street legal. The race cars are not street legal and are built for a very specific reason (to go around tracks as fast as possible). This is similar to your on premise Data Center compared to a cloud Data Center.

To use a cloud Data Center better or to its fullest potential, using it efficiently is a way to control Costs. Efficiency can be done by checking if your current resources in the cloud are used efficiently and combining resources to lower your costs.

Shut down unused resources

Cloud Clean-up can be done in the number of vendors (from many to one or two), but also in the number of cloud instances, which are already running in your environment. Through automation, organizations can simplify processes and save time-consuming manual operations.

The easiest way to optimize cloud costs is to look for unused or unattached resources. Shutting down unused instances, especially during weekends and at the end of the workday, is the first step in optimizing costs. The recommended actions are to shut down or resize, specific to the resource being evaluated.

Spot instances

Cloud vendors like AWS, Google and Microsoft have services where you can request unused instances to reduce compute costs and improve application throughput.

Spot instances enable an organization to deploy a broader variety of workloads, while enjoying access to discounted pricing compared to pay-as-you-go rates. Spot VMs offer the same characteristics as a pay-as-you-go virtual machine, the differences being pricing and evictions. Spot VMs can be evicted at any time if the cloud vendor needs the capacity.

Workloads that are ideally suited to run on Spot VMs include but are not necessarily limited to the following:

• Batch jobs
• Workloads that can sustain or recover from interruptions
• Development and test
• Stateless applications that can use Spot VMs to scale out, opportunistically saving cost
• Short lived jobs which can easily be run again if the VM is evicted.

Spot instances are an excellent way to reduce cloud costs, but they can be complex to manage.

Conclusion on Clean-up

Everybody loves a good clean-up, whether it is your desk, your house or the cloud environment. A good and thorough clean-up works productively and, when done regularly, cuts down costs.

Efficient usage and shutting down of unused resources sound simple and logical, but it is mostly overlooked at or simply not done. Using Spot instances looks appealing from the cost aspect, but there is an important caveat; reliability is not guaranteed. The cloud provider can interrupt these instances at short notice to reclaim capacity. However, with careful management, spot instances can be useful for batch processing and high-performance computing (HPC) clusters, web server clusters, and many other workloads.

-> Are you looking for help in executing any of our 5(+1) ways? Let me know in the comments or send me a DM. <-

Part 1 -> Automation