Consumption Analytics Documentation

Home > Documentation for older versions > Cloud Cruiser 3 > Setting Up Collection > Native collectors > Microsoft Azure Services for Windows Server > Architecture and data flow

Architecture and data flow

Table of contents
No headers

The following diagram shows the architecture for this solution:

Architecture for collection and processing of Azure Services for Windows Server data


Cloud Cruiser’s native Azure Services Usage Collector, which you run in a daily batch job, queries the Azure Services for Windows Server Usage API for all usage data since the last time it ran. The collector receives this data as JSON documents, which it writes to a feed directory in the <working_dir>/usage_files directory. Separately, Azure Services for Windows Server uses the Cloud Cruiser REST API to write data about offer and subscription events to subdirectories under that feed directory, also as JSON files.

A separate daily job runs afterward to process this data and load charges. It uses the JSON Collector to convert both the usage data and the event data from JSON format to CC Records. It uses a feed of state CC Records to track resources whose create and delete events don’t fall within the same day, keeping them allocated until their delete event arrives.

The job then uses a translation table built from subscription records to look up the subscriber who corresponds to the subscription ID in each usage record, adding that subscriber to the record as an identifier. For all subscribers that do not exist as customers in Cloud Cruiser, the job imports them as new customers. It then transforms subscriber names into account IDs in the usage data and loads charges into the database.

Each run of the Azure Services Usage Collector ends by updating a file in its feed directory with the ID of the last record it processed. Each subsequent run of the job will collect only usage records created after that record.

Last modified


This page has no custom tags.


This page has no classifications.

 (c) Copyright 2017-2020 Hewlett Packard Enterprise Development LP