Client: One of the UK's largest financial institutions
FinOps within a Multi-Cloud EnvironmentÂ
Read More
Tailored cloud financial management solutions.
Manage the end-to-end delivery of IT services to customers successfully.
Complete cloud migration services for the public sector.
Transforming the financial services sector with industry-leading cloud and data solutions.
Delivering compliant and secure cloud automation solutions for the Public sector.
Other sectors that we currently have clients within include insurance, media, retail, construction and automotive.
Our team of industry-leading experts deliver world-leading transformation solutions.
We would love to talk about transforming your business. Please let us know.
Our latest industry news, insights, employee stories, and upcoming events.
2023-06-05 00:00:00
I was sitting at my desk, trying to figure out the simplest way to describe how to implement some of the FinOps phases for a client presentation, when I received an email containing our internal AWS bill.
Our AWS accounts are mainly used for educational purposes, to create sandboxes and to play with the latest cloud services and DevOps tools, I assumed that the monthly cost wouldn't be too high, so imagine my surprise when I discovered that the bill amounted to $2000! Given the size of our company and cloud estate, I was imagining something closer to $500.
My inner FinOps mindset immediately kicked in, and I found myself scrolling the 98 (!) pages of the bill, looking for the details of spend per AWS service, and, unsurprisingly, I discovered that at the top of the list, were some of the usual culprits.
Service | Monthly Cost $ |
---|---|
Amazon Elastic Compute Cloud | 561.35 |
Elastic Load Balancing | 231.56 |
Amazon OpenSearch Service | 146.20 |
Amazon CloudWatch | 61.26 |
I kept scrolling and found the cost allocation per each linked AWS account, and besides one account hosting our solution for cloud observability and compliance, Airwalk Airview, the rest were mainly sandboxes, but some were really expensive ones, so I reached out to the manager who allocated the accounts to discuss further and to identify who was responsible for each.
As a further exercise I identified the possible ways to reduce the cost of such accounts and based on the cloud services described above I produced actions and strategies, I also classified them into tiers by effort and complexity.
Service | Low hanging fruit | Intermediate | Advanced |
Amazon Elastic Compute Cloud |
|
|
|
Elastic Load Balancing | Remove unused instances |
|
Use S3/CDN to store static content and reduce traffic |
|
|
Move to serverless | |
Amazon CloudWatch | Turn off detailed monitoring for non-production | Reduce the number of alerts, metrics and dashboards Reduce log retention |
Migrate to a centralised observability layer to provide more value and reduce cost |
In the end, I reviewed my analysis and realised that I had quickly implemented two out of the three canonical FinOps phases: Inform and Optimise.
The above was only a theoretical exercise, but there is a very important message that can be distilled from it. The FinOps mindset can be adopted using really simple steps and on a small scale, and then its level of maturity can grow, in line with your company's journey to the public cloud.
Are you looking for your own way to implement FinOps? We can help, get in touch with us.