Renewable Record Breakdown - rollthecloudinc/hedge GitHub Wiki

This document will discuss the purpose of each field inside the renewable record. The renewable record is generated added to Open Search every lambda execution. Additionally, a record for each region is created to provide comparison between regions on future charts. Given this data we can determine how much impact bouncing API requests between regions with lowest grid intensity actually has on reducing carbon emissions. We can compare the emissions of a service sitting in one region to that of a service/lambda that has been bounced between several regions. This will fill the existing measurement gap of hedge for internal services and octostore. Future additions will allow services to send back these metrics within response headers to integrate into the system using their own algorithms for calculations.

This strategy is inspired by Cloud Carbon Footprint. However, Cloud Carbon Footprint lacks real time calculations, actual memory part of energy calculation, near realtime grid intensity part of carbon calculation and open search integration + serverless solution. Therefore, the decision was made to integrate this directly into emissionless rather than using Cloud Carbon Footprint. Once integrated with Open search dashboards can be created that don't require any additional cloud servers / resources beyond Opensearch nodes and CDN to host web experience / dashboard.

Field	Description	Unit
Id	Unique GUID
RequestId	Unique request ID
Region	HEDGE region
AwsRegion	AWS Region
Duration	How long the lambda ran	ms
BilledDuration	The amount of time billed for lambda execution	ms
MemorySize	Amount of memory allocated for function.	mb
MaxMemoryUsed	Maximum of memory the function execution actually used	mb
InitDuration	Amount of time lambda to initially start-up	ms
Intensity	Energy grid intensity inside executed region	g/kWh
Electricity	The amount of electricity used to execute lambda	kWh
Carbon	The amount of carbon emitted by function execution with respect to duration, intensity, memory	@todo
Function	The function name that was executed
Path	When executed as part of api request the url path
Called	Whether this is the called record or a placeholder for cross region carbon production comparing
Organization	The organization that executed the function
Repository	The repository that was used during function execution.
Service	The HEDGE service name
Resource	When executed through api gateway the resource path

type RenewableRecord struct {
	Id             string  `json:"id"`
	RequestId      string  `json:"request_id"`
	AwsRegion      string  `json:"aws_region"`
	Region         string  `json:"region"`
	Duration       uint16  `json:"duration"`
	BilledDuration uint16  `json:"billed_duration"`
	MemorySize     uint16  `json:"memory_size"`
	MaxMemoryUsed  uint16  `json:"max_memory_used"`
	InitDuration   uint16  `json:"init_duration"`
	Intensity      uint16  `json:"intensity"`
	Electricity    float64 `json:"electricity"`
	Carbon         float64 `json:"carbon"`
	Function       string  `json:"function"`
	Path           string  `json:"path"`
	Called         bool    `json:"called"`
	Organization   string  `json:"organization"`
	Repository     string  `json:"repository"`
	Service        string  `json:"service"`
	Resource       string  `json:"resource"`
}