Managing Costs in the Cloud Part 2
JohnCrawfordIBM 100000BANX Visits (1914)
Managing Costs in the Cloud
By Paul Quigley
This is the second of a two-part blog focusing on implementing IBM SmartCloud Orchestrator 2.3 with IBM SmartCloud Cost Management to help you measure cloud resource usage and accurately charge your customers for their cloud resource usage. Part 1 focused on the architecture and products involved. Part 2 will focus on the cloud service usage data collected and the reports that are available.
What type of cloud service usage data is collected?
SmartCloud Orchestrator Enterprise produces cloud service usage data per virtual machine, including:
How is the cloud service usage data collected?
The diagram below provides a high-level overview of the data collectors, rate templates, cloud reports, and how they are related.
Click on iImage to view in separate window
SmartCloud Orchestrator Enterprise provides four job files to collect data across three different data collectors, as follows:
Each data collector collects specific data on an hourly basis, logging the data to files with common source resource (CSR) records.
Data from the CSR files is merged and correlated in SmartCloud Cost Management by running jobs to process the data. By default, this occurs at 21:30 each day. The new (merged) file is then used for analysis, reporting, and billing.
Merging the CSR files allows you to see reports based on the tenancy data. For example, customer “A” can only see their cloud service usage data and customer “B” can only see theirs. Invoices can be generated using different rate structures for each customer.
When a report is requested, SmartCloud Cost Management will apply the appropriate rate tables and tenancy data, as you will see in the examples later in this blog.
Defining rates for chargeback
SmartCloud Cost Management supports several rate templates for cloud resources, such as:
Hosting charges, such as:
Once the rates are applied against the cloud service usage data, you have a chargeback mechanism and can generate an invoice to send to your customers.
Tip: The cloud service usage data tells you how long cloud resources have been reserved, not necessarily in use. To more accurately reflect actual cloud usage, you can use the infrastructure charges as a base cost and define additional costs based on license, operating system, size of VM, and so on, using the license, hosting, and sized-hosting rate templates.
Cloud costing example
Using the default license charges rates, suppose you provision a virtual machine with Windows Server 2012, DB2 v10, and WebSphere Application Server v8.
For every hour that the virtual machine is provisioned, the cost would be:
Let’s look at an example invoice. In this case, the customer is Customer P.
Click on image to view in separate window
Notice, for example, the infrastructure charges of $353.60 for their x86 platform. The cost is further broken down by CPU, memory, and disk, as well. The invoice also includes additional costs for software licensing and hosting charges.
Example usage report
Let’s look at an example usage report. In this case, this is a Project Summary report looking at all virtual machines across all projects. The bar graph shows the provisioned virtual machines from 0 through 1000 days.
Click on iImage to view in separate window
The data has been filtered for only the virtual machines that have been deployed in the last 51 to 100 days. There are approximately 20 virtual machines that have been deployed. All virtual machines are active. Looking at the pie chart, you can see that the virtual machines have been deployed evenly across several projects. The table at the bottom of the report provides details regarding each virtual machine.
SmartCloud Orchestrator creates cloud service usage and tenancy data as it provisions virtual machines. Several data collectors are used to collect the virtual machine resource data as well as tenancy data.
With SmartCloud Cost Management you can understand your costs and track, allocate, and invoice based on actual resource use by department, user, and many additional criteria. Reports can be run to generate invoices (bills) based on customer usage of cloud resources.
For more information
For more information about IBM SmartCloud Orchestrator, visit
About the Author
Paul Quigley is a Senior Technical Enablement Specialist with the IBM Cloud and Smarter Infrastructure Enablement team, with expertise in Tivoli Service Automation Manager, Tivoli Provisioning Manager, and Smart Cloud Orchestrator.