I know what you are thinking, so often these posts are filled with either overly simplified common-sense advice or conversely, an elaborate pipedream which only works in a perfect world. Neither of which are of any value but are often presented as the most important information ever. That is not the intention of this post, and to save money in the cloud is to follow the same advice we were given by our parents for years.
“Turn the lights off when you leave the room.”
It really is that simple! However, while you can see when the lights are on, you often don’t know when you are consuming cloud resources (or how many). It is not uncommon for a large organisation to have hundreds or thousands of workloads running either when there is no staff actively working (outside of hours of operation), or when they have moved on to a different project and no longer need that specific workload. Imagine a shed with the lights on but no idea where that shed is located.
As a real-world example there were two staff who had identical jobs in an organisation. They were responsible for managing their own Microsoft Azure accounts (the company would reimburse them). One was incredibly diligent and had approximately $400-$500 per month in cloud spend. The other was probably as cavalier as one could be. Not only did they not turn off workloads when they were finished for the day, they would often spin up entire new workloads. As you could imagine, their bills grew exponentially to the point where every four to six months, they would do a one-off clean up. Over the six months, these disparate workloads cost the company over $23,500 in cloud spend.
It is a simple problem with a simple solution. However, with the easy to consume model of cloud, it can be very difficult to manage for many organisations. Some options are as follows:
- Change behaviour of the end users who consume cloud. There are several challenges with the approach, namely, it may be prone to errors and is a very manual process. If an organisation were to take this approach, it would require a cultural shift and it might still not be effective.
- Hire someone (or several people) to apply organisation-wide governance. This can be achieved in several ways but specifically, you would require someone to make sure the appropriate scripts were deployed. They would also have to hold those personnel accountable. This method can cause issues because it does not scale and has many reactive elements to ensure success.
Now, I understand that this is all straight forward, and I promised that this would be a practical guide to saving money in a cloud environment. Well Buttonwood is a Cloud Management Platform and specialises in financial governance in multi cloud environments.
Historically, it is easy to apply new technology and procedures within Greenfield environments (where workloads being deployed are new), however, there are a range of challenges you experience when trying to make the same changes in Brownfield environments (where resources have already been deployed and now need to be managed). We knew that to truly allow any organisation to save money on their cloud spend, we would need to apply the same financial governance on all workloads – whether new or existing.
Buttonwood provides a simplified approach to putting existing workloads under management and start saving money today.
Here is how:
- Start up Buttonwood Automate and navigate to the Brownfields module to create a new project
2. Provide a name and description for the project:
3. Select what Environment this project will inherit its governance policies from as well as the account to link resources from:
4. Select the network which contains the resources which are to be linked to the project
5. And finally, create the project
6. Now that the project is created, resources can be linked.
Drag and drop as many compute nodes as required and link them to a resource. In this case, we are linking to an Amazon Web Services (AWS) instance, specifying its Operating System type, and providing credentials for how to connect to the resource
7. We can now deploy this project:
8. As part of Buttonwood’s financial governance requirements, deployments require a Cost Centre where resource consumption costs will be allocated, and an expected lease period to ensure that unused services don’t run indefinitely (this lease can be changed as required during the lifetime of the resources).
9. We can see that this deployment will inherit the Hours of Operation policy from the selected environment. Being a Development environment, the policy has been configured to automatically power on the resources during weekdays at 7am, and then power them off again at 7pm, saving hourly running costs often charged by service providers.
In some cases, you might be bringing in multiple resources which have order dependencies for power on and power off – this can be managed once the deployment is up and running.
10. Once deployed, the mapped resources are now considered “under management”, inheriting governance and automation features such and lease management, cost monitoring, power sequencing, and other Day 2 operations.
This simple, automated, process alone can save 65% on your organisation’s cloud bills. It will ensure that when your staff go home, their workloads are automatically powered down in a consistent manner.
I know that seems like a lot to take in but, within minutes you can apply policies on your existing workloads. Don’t take our word for it. You can register for a free trial account and start saving money today.
You will find the guide for deploying Buttonwood Automate here.