Lower Cloud costs by reducing the size of systems

 

Cloud costs : #1 concern for CIOs 

Downsizing systems – the right option

A recent report from Flexera (end of 2022) indicates, and for the 7th  consecutive year, that optimizing the use of the Cloud remains the main subject of concern for CIOs. And yet, relatively few companies are moving forward with real continuous optimization projects. Which, however, seems quite obvious.

1.     The growth of costs in the Cloud, a shared observation:

A 2022 Gartner survey indicates that companies expect a 29% increase in Cloud spending in 2023 compared to the current year. 

 

 

This same study indicates that 2023 costs will be 12% higher than planned costs.

Objectively, no one can dispute that Cloud costs look like a headlong rush. There are always more projects, and they over-consume data.

For example, Netflix spends nearly $10 million per month to store/process its data on AWS.  Some of our clients are experiencing this same meteoric growth.  

So much so that some companies are turning back and moving on to their own infrastructure.

Hyperscalers are well aware of the risks weighing on them: they set up accounting tools and alerts that are triggered before invoices reach stratospheric levels: Azure Cost Management, Google Cloud Cost Management, and AWS Cloud Management Financial for Hyperscalers.

We also think of Data Dog which carries out these analyzes on multi-Cloud architectures.   Densify, another player in the space, focuses on cloud resources such as instances, Kubernetes clusters and VMware. We could also cite Flexera which offers multi-cloud accounting for expense tracking with elaborate reports broken down by team and project. Flexera One also offers methods to optimize consumption by targeting unnecessary allocations.

These tools are expensive, and they do not make it possible to simplify Information Systems.

2.     The carbon impact of Cloud Information Systems is growing:

The physical consequences of digital activities are increasingly difficult to ignore. Digital technologies are responsible for 4% of global greenhouse gas emissions and their energy consumption increases by 9% each year. Among these emissions, around 15% are due to the energy consumption of data centers .  So it can be useful to put the brakes on the inflation of data and its processing, especially when the data is massively replicated, obsolete, useless, etc.  

3.     The intelligibility of systems is challenged:

According to an Oracle study, "The Decision-Making Dilemma", in which more than 14,000 IT professionals in 17 countries were surveyed, almost eight in ten respondents (78%) believe that they have never had so much data from different sources available to support their decisions. But for 86% of respondents, this volume makes decision-making far too complicated. These difficulties in making decisions lead to negative repercussions for 85% of those surveyed: peaks of stress (36%), missed opportunities (33%) or unnecessary expenses (29%). A real problem, and we are not getting better. 

4.     The methodology we propose to change the paradigm:

Around information systems databases gravitate audit databases which include, among other things, all the uses that are made of the data of a system, and in contrast the data which are not subject to 'no use.

Starting from unused data at the end of the flows, ie from these "informational dead ends", our {openAudit} software continuously identifies all of the upstream "dead branches", automatically going back to the operational sources.

• These will be components that store data (tables, files, dashboards);
• It will also involve all the code that allows the data to be processed in the Information System.

 

For what ? because 80% of the costs of a Cloud system comes from these 2 positions: storage / processing.

Cloud Cost Allocation

 

Thus, it becomes possible to effectively and permanently rationalize an Information System. Our projections tell us that 50% of the systems' storage/processing can be decommissioned, without the services provided being degraded . 

 

 

Conclusion 

Sobriety is a fashionable term. For excellent reasons. It must apply to Cloud Information Systems which have become insatiable, with countless side effects. 

Ellipsys is able with its {openAudit} software to identify “dead matter” in systems and continuously – this dead matter can be removed so that the systems remain as simple and coherent as possible, but also to reduce energy consumption and GHGs.

An essential virtuous spiral.

 #ellipsys #datalineage #ITdebt #greenops #finops

www.ellipsys-lab.com