Commercial Director Lota.cloud
Controlled cloud spending for reduced carbon impact
Controlling, understanding and optimising your Multicloud expenditure towards a minimised carbon impact with Lota.cloud
The promise of the cloud, from a financial point of view, is a reduced infrastructure cost compared to an on premise IT infrastructure.
Lifting & shifting your architecture to the Cloud as it is is only a technological or temporal choice but certainly a bad financial operation in the management of your CSP expenses.
Pay as you go will only be interesting if the migration is based on an architecture redesigned for the cloud and with a view to FinOps by design.
The complexity of the offer, the opacity sometimes present in the pricing of hyperscalers does not facilitate the task. Many companies only become aware of the waste at the end of their migration, after several months or years.
While it is relevant to make savings by purchasing the desired technology at the best price, this centralised aspect of FinOps around purchasing is insufficient for good management and to aim for a reduced carbon impact of the products and services associated with the infrastructure.
Multicloud, various services and products in their categories and variations = complexity for the visibility of expenses
It is now common knowledge that wastage on cloud spending has reached one third of the bill on average, with an annual acceleration of 10% each year (30% wastage by 2020 known benchmark).
It is therefore equally important to realise that each IT department, each production, development, research or test department generates its own waste in the use of resources and therefore in its billing.
In the context of the accountability of the players, the practice of showback is an essential step in deducing the good or bad use of resources from the sharing of cost visualisation. To this end, a good tagging policy and a multicloud dashboard aggregating billing information will enable the first obligation of a Finops approach to be met: information.
Tagging will identify the allocation of resources by aggregating them according to its cloud architecture, and your dashboard will show the coverage rate and financial impact of untagged resources, as well as the progress in the tagging process.
This shows the importance of having these indicators available at all times and in real time in the CSR support of a company's IT organisation.
The more your resources are under control, the more you will be able to measure waste and act directly on the carbon impact of your public cloud consumption. Action will be all the more effective if each stakeholder is involved in the FinOps approach according to their contribution to usage.
It should be remembered that the classic scheme of reserving virtual instances will allow significant savings but could have a negative effect on the carbon consumption of the infrastructure. Indeed, this centralised approach sometimes results in a useless multiplication of the machines used in view of the comfort given and the reduced price.
It is often more beneficial to choose newer, more efficient machines that offer a better price/performance/volume ratio.
Among the recommendations related to FinOps best practices, the management of snapshots, the hitching of Dev instances on weekends, the use of Kubernetes and Severless are optimization paths that favourably influence savings and therefore the minimization of the carbon impact.
On the infrastructure side of the CSPs, a notable effort has been made to offer water and energy efficient data centres with adiabatic cooling to support the public cloud.
They are supplied with the latest generation of servers that operate at a higher optimum temperature, which greatly limits the need for air conditioning. Another way of doing this is by water-cooling the equipment in the rack.
All of this makes it possible to respect commitments in line with Green IT, and therefore to support companies in a more reasoned Cloud policy, and complements GreenOps, which extends a FinOps approach.