Services

An introduction to Services

In Exivity Services can be anything that corresponds to an SKU or sellable item from your Service Catalogue. It should relate to a consumption record (or multiple records) from your extracted data sources.

For example: with most public cloud providers, the provider defines the chargeable items that are shown on the end of month invoice. However, when working through a Managed Services Provider, a Cloud Services Provider, or a System Integrator, additional services can be sold on top of those. Potentially, you may want to apply an uplift to the rate or charge a fixed amount of money every month for a certain service. Different scenarios are possible here, it all depends on your business logic.

Terminology

A service is a named item with associated rates and/or costs used to calculate a charge that appears on a report, where rates represent revenue and costs represent overheads.

When discussing services and their related charges several terms are required. Exivity uses the following terminology in this regard:

Creating service definitions

When created during the ETL process, service definitions are created via the services statement in Transcript. During the execution of a Transcript task, service definitions created by these statements are cached in memory. Once the task has been completed successfully, the cached services are written to the global database where they remain indefinitely (or until they are manually deleted).

If the task does not complete successfully then the service definitions cached in memory are discarded, the expectation being that the task will be re-run after the error condition that caused it to fail has been rectified and the services will be written to the global database at that time.

Types of charges

There are different types of charge that can be associated with a service. Collectively these influence the total charge(s) shown on the report and Exivity supports the following charge types as described in the Terminology table above:

  • unit rate

  • COGS rate

At least one of these charge types must be associated with a service definition.

Once the resulting charge has been calculated based on the charge types, it may be further modified through the application of adjustments, proration and minimum commit (all of which are detailed later in this article).

Charge intervals

In order to calculate the charge(s) associated with usage of a service Exivity needs to know the period for which each payment is valid. For example, a Virtual Machine may have a daily cost associated with it, in which case using it multiple times in a single day counts as a single unit of consumption whereas Network Bandwidth may be chargeable per Gigabyte and each gigabyte transferred is charged as it occurs.

The charge interval (also termed simply interval) for a service can be one of the following:

  • individually - the charge for a service is applied every time a unit of the service is consumed, with no regard for a charging interval

  • daily - the charge is applied once per day

  • monthly - the charge is applied once per calendar month

Although hourly charge intervals are not currently supported directly, it is possible to charge per hour by aggregating hourly records and using the EXIVITY_AGGR_COUNT column created during the process to determine the units of hourly consumption as a result.

Charge models

Monthly services may be charged in different ways:

Peak

For each day of the month, a 'candidate' charge is calculated using Quantity * Unit Rate. The monthly charge will reflect the day of the month which resulted in the highest charge.

If multiple days share the same highest charge then that charge will be associated with the first of those days seen, unless a subsequent day in that set has a higher quantity, in which case the charge will be associated with that subsequent day.

Average

The average unit rate for those days where usage was seen in the month is calculated and multiplied by the average quantity for each day in the month. When calculating the average quantity, any days for which there was no consumption are factored in as having a quantity of 0.

If Average Charging is applied in combination with proration, then the resulting average unit price as shown on reports may be less than you expect to see. This is because the average unit price as shown on the reports is calculated using charge / average_quantity and proration will reduce the charge if there was no consumption on one or more days of the month, resulting in a lower average unit price.

Specific Day

The charge is based on the quantity consumed on a specific day of the month.

Last Day

The charge is based on the quantity consumed on the last day of the month.

Minimum commit

The minimum commit is the minimum number of units of consumption that are charged every interval, or (in the case of services with an interval of individually) every time the service is used. If fewer units than the minimum commit are actually consumed then the service will be charged as if the minimum commit number of units had been used.

Proration

After the charge for usage of a monthly service has been determined, it may be prorated by modifying that charge based on the frequency of the usage.

This process will reduce the charge based on the number of days within the month that the service was used. For example, if consumption of a service with a monthly charge interval was only seen for 15 days within a 30 day calendar month then the final charge will be 1/2 of the monthly charge.

Service definitions

A service definition comprises two categories of information:

  1. The service - Metadata describing fixed attributes of the service such as its name, description, group, interval, proration and charge type(s)

  2. The rate revision - Information detailing the charge type(s) associated with the service (the rate and COGS values) and additional information detailing the date(s) for which those values should be applied

A service definition is associated with a specific DSET as the units of consumption are retrieved from a column (named in the service definition itself) in usage data.

The following tables summarize the data members that comprise each of these categories:

Service attributes

RDF, rate_type and cogs_type are automatically derived from the parameters provided to the service and services statements

Rate revision attributes

The rate_col and cogs_col are used when the specific value to use is derived at report-time from the usage data, as opposed to explicitly being included in the rate revision itself.

A service may have any number of associated rate revisions so long as they have different effective_date or minimum commit values. This means that a service can have different charges applied depending on the date that the report is to be generated for, or depending on the specific values in the columns used by a report.

A service may use either or both of rate and cogs.

Either of rate or cogs may have a value of 0.0, in which case no charges will be leveraged against the service but the units of consumption will still be shown on reports.

Last updated