This new capability allows you to collect usage information from SCP environments using the SCP High scale low touch (HSLT) commands. The new HSLT SCP Collector gathers usage data every hour and processes it once a day. Usage, Detail and Identifier data is stored on a daily basis. The usage data is then billed, stored and can be reported on on a monthly basis.
A sample job file is provided as part of this functionality to show how to bill each access-id for the high-water mark of allocated resources in the month. The sample job file, SampleHSLT_SCP.xml is divided into three separate jobs.
The first job, SCP_collect_HSLT_hourly_data is recommended to be run every hour at XX:59. This job will run HSLT commands to collect all relevant resources for each access id that is using the SmartCloud Provisioning Service. Firstly, a list of all available access ids is collected using the command iaas-describe-accesses-by-user.
Then, for each access id, the command iaas-describe-resources-inuse-by-access is run to collect the relevant resources for that access id. The resources gathered per access id include:
Memory (MB) , Volume (GB), Number of Virtual Processors, Number of VM Instances, and Number of static IP Addresses.
The HSLT commands also provide context information that feeds into the Account Code Structure. The Account Code Structure includes the following identifiers:
- Service Region
- Group, Username
- Access Id
The second job, SCP_Process_daily_data is recommended to be run every day some time after midnight. This job will process the daily CSR file and extract the maximum value across the day for each resource for each access id. The resource values are then stored in the cimsresourceutilization table of the SmartCloud Cost Management database. Detail and Identifier data is stored in the cimsdetail and cimsident tables of the SmartCloud Cost Management database.
The third job, SCP_Process_monthly_data is recommended to be run once a month at the start of the month. It will process the last months worth of data from the cimsresourceutilization table. It will do this by extracting the maximum value for each resource for each access id. Billing is applied to the data using the relevant SmartCloud Cost Management rate codes and the processed data is then stored in the cimssummary table of the SmartCloud Cost Management database, allowing reports to be run on the data.
The sample jobs can be customized for other charging algorithms if desired. Examples include charging on a daily (or hourly) basis in addition to or instead of on a monthly basis. Tiered pricing logic can be applied to provide for having charging amnesty for users/departments that stay below a certain threshold.
Rates are defined for each resource. These rates are used for billing purposes.
Additions have also been made to the existing SmartCloud Cost Management KVM collector to include new resources and a separate job file has been included to to add some SCP context data to the Account Code Structure, achieved by running HSLT commands.
For information about the existing TUAM KVM collector refer to the following link in the TUAM 7.3 Information Center:
The new resources for the KVM Collector include Bytes Received, Packets Received, Receive Packets dropped, Receive Packet errors, Bytes Transferred, Packtes Transferred, Transfer Packets dropped, Transfer Packet errors, Log Size of VM Image, Size of VM Image on Disk.
The new Account Code Structure for the KVM Collector contains the following identifiers: Service Region, Group, Username, Access id, VM Name
The VM Name contains the Access id allowing the information collected from the Hypervisor to be related back to the SmartCloud Provisioning identifiers.
The following reports are sample reports run on a system that has collected data from one Service Region on a SmartCloud Provisioning System:
- Top 10 Pie Chart
- Invoice by Account level
- Run Total Invoice
See the Information Centre (http://pic.dhe.ibm.com/infocenter/tivihelp/v3r1/topic/com.ibm.ituam.doc_7.3/admin_win_dc/c_core_data_collectors.html) for details.