Data aggregation

Data aggregation is a method of accumulating data into an accounting record that is otherwise presented through multiple records.

Aggregated data is written periodically, according to the intervals described in Interval accounting.

Accounting data is totaled inside the kernel with no impact to applications or middleware. The data is made available by interval accounting, which is a kernel function that periodically writes these records to the acct file. When the kernel aggregates records, it maps them to a set of aggregated records that are managed internally. These records are pending in the sense that they have been input to the system, but have not been committed to the recording mechanism inside Advanced Accounting. Interval accounting is used to commit the aggregated records so that they are written to the acct file.

Because aggregated data is recorded using different data structures, you must verify that the billing application recognizes these structures. Consult the documentation provided with your billing application to determine if your billing application supports data aggregation. Data aggregation can be enabled or disabled at the system level or project level.

System-level data aggregation commands and fast paths

You can manage system-level data aggregation using the command line or SMIT fast paths.

System-level data aggregation commands allow you to turn on and off process data aggregation, kernel extension data aggregation, and ARM transaction data aggregation. The following table lists the commands you can use to turn on and off process data aggregation, kernel extension data aggregation, and ARM transaction data aggregation.

Table 1. System-level data aggregation commands

Task	Command	SMIT fast path
Enable or disable system-wide process aggregation.	acctctl agproc {on|off}	smit system_paggr
Enable or disable system-wide aggregation for third-party kernel extensions.	acctctl agke {on|off}	smit system_kaggr
Enable or disable system-wide aggregation for ARM transactions.	acctctl agarm {on|off}	smit system_aaggr
Query the overall accounting state	acctctl	None

Project-level data aggregation commands

You can manage project-level data aggregation using the command line or SMIT fast paths.

The following table shows the commands to turn on and off aggregation for individual projects.

Table 2. Project-level data aggregation commands

Task	Command	SMIT fast path
Enable or disable aggregation for a project, specified by the projname parameter.	projctl chattr agg projname {-s|-u} [-d projpath]	None
Query the aggregation state for all projects.	projctl qprojs [-n]	None
Query the aggregation state for the specified project.	projctl qproj [projectname]	None