COST

The true key to understanding total cloud cost of ownership is built upon Transparency and Accuracy. In order to add Transparency it is recommended:

• To segment and allocate costs to each business department; the team; or the project. The majority of cloud costs can be categorised and directly allocated to an organisational unit.
• Shared cloud costs should be identified upfront as part of system design. Shared costs can be split using an appropriate distribution model(s) (proportional, fixed, even-split) across the entirety of the organisation.
• For shared platforms and accounts leverage a label (tag) strategy for cost allocation. Allocate cluster costs in proportion of each cost centre label’s usage.

The process of minimising waste of Cloud resources by finding the right Resource Size and Instance type for different applications and Use-Cases. It is best to scale out in the Cloud and Autoscaling should be used to handle increases in load.

Auto-scaling is a way to automatically scale up/down or in/out the number of compute resources that are being allocated to your application based on its needs at any given time.

Pay-as-you-go is a payment method that charges based on usage. You only have to pay for a particular resource or service for as long as you need it, which prevents the waste of resources.

A Spot Instance is an instance that uses spare Cloud Provider Capacity that is available for less than the On-Demand price. The discount might be as large as 70-80% from the regular price. However, Spot instances can be terminated by the Cloud provider at any time with a very short notice.

A Reserved Instance is an instance that is prepaid for a longer commitment period of 1 to 3 years. The discount can be in a range of 30-50% from the regular on-demand pricing, but you’ll have to pay for it even if it is not needed anylonger.

Scaling out involves adding more computational units to a system in order to add capacity (e.g. adding 5 VMs into a pool), scaling up involves resizing existing units by adding resources into them (e.g. using larger compute flavour).

It all starts with Storage, Compute and Network. In general, cloud costs are calculated based on Performance and Consumption values. Most Cloud providers charge at least for the following:

• Compute VMs running time
• Allocated block storage space
• Object storage by usage and number of requests Network traffic both within the region and out of the region
• Capacity of Load Balancers based on number of requests and connections

Serverless lets you run code without provisioning or managing servers. It can run code for virtually any type of application with zero administration and only pay for what you use. You are charged for the number of requests for your functions and the duration it takes for your code to execute. Duration is calculated from the time your code begins executing until it returns or otherwise terminates, rounded up to the nearest 1 ms. The price also depends on the amount of memory you allocate to your function and most providers allow you to allocate RAM with MB precision.

Lock-in occurs when a customer must continue to use a particular cloud provider because its proprietary services, APIs, or protocols make it difficult to move or migrate a workload, making customers more dependent (locked-in) on a single cloud provider.