Configuring aging properties of systems and business drivers

Use the Aging configuration page to set parameters that impact the efficiency and the disk space needed by the data warehouse. For each table that stores TrueSight Capacity Optimization metrics, you can set for how long the data must be kept and whether to partition the table into smaller chunks, if needed.

You can dismiss and purge old and unused data that is associated with systems and business drivers. For more information, see Enabling automatic dismissal and purging of systems and business drivers.

To access the Aging configuration page, under the Administration tab, click Data Warehouse and select Aging configuration

For more information, see the following sections:


To view and manage aging and partitioning configuration details

Click Advanced aging configuration to view a page that displays a summary of the aging and partitioning configuration for each data table, including:

Example of the Custom tables aging configuration table

Click on the name of a table to view the detail page for that configuration, where you can edit or delete its settings.

Example of detail page for the SYS_CONF_STAGE table

To enable or disable aging configuration for the data warehouse

In the Aging configuration page, use the Data Warehouse aging state option to enable or disable aging configuration for the data warehouse.

  • Click to enable the aging configuration.You must ensure that the historical data import is complete before you enable aging.
  • Click to disable the aging configuration. You must disable the Data Warehouse aging configuration before you start a historical data import.

To configure aging parameter for metrics

Click Advanced aging configuration to configure the Aging parameter. This parameter defines the number of days that the specific metric value will be kept in the data table, using its timestamp as starting date. It is important to properly tune this parameter, as it can strongly influence the space occupied in each tablespace and the availability of the data for analysis purposes. Following type of tables are available in the aging configuration:

  • *_CONF_STAGE: these tables are used to temporarily load the ETL data until the data warehouse copies it to the DETAIL tables

  • *_DATA_DETAIL: these tables are used to store more detailed resolutions

  • *_DATA_DETAIL_SPLIT: these are internal tables used by near real time warehouse and are not used in analyses, models, etc.

  • *_DATA_DH: these tables are used to store data at hour resolution

  • *_DATA_D: these tables are used to store data at day resolution

  • *_DATA_MDCH: these tables are used to store data for month at day resolution

  • *_DATA_STAT: these tables are used to store system statistics

For example, consider the table SYS_DATA_DETAIL. It holds the system metrics collected at the maximum detail available, given that the duration of the samples is greater or equal to 5 minutes. According to your needs, you can decide to reduce or increase the aging parameter value for this table:

  • A small value will save space because the detailed data will be deleted after a short period of time, but will prevent you from performing detailed analyses
  • A high value will let you perform detailed analyses over more data, but the tradeoff will be a high space consumption and slower performances

The short, medium, and long properties are used in the aging class mappings. Metric profiles, when used in combination with the aging class mappings, enable you to be more granular with the metrics that are included or excluded for a specific ETL. These metric profiles also define the duration for which the data is retained in the data warehouse. This aging configuration affects the amount of storage that is used by the data warehouse.


DetailHourDayMonth

SystemBusinessSystemBusinessSystemBusinessSystemBusiness
ShortOne weekOne monthTwo monthsOne yearFour monthsTwo yearsOne yearTen years
MediumTwo weeksOne monthFour monthsOne yearEight monthsTwo yearsTwo yearsTen years
LongOne monthOne monthSix monthsOne yearOne yearTwo yearsThree yearsTen years


The Aging parameter also regulates the data retention period of BMC TrueSight Capacity Optimization. As data is automatically aggregated at different time levels, when the highest level data (usually, WKLD_DATA_MDCH and SYS_DATA_MDCH) becomes old (i.e. its aging period expires) it is deleted and all information about that period of time is lost.

Moreover, the Aging parameter directly affects the maximum age of the data that can be imported in BMC TrueSight Capacity Optimization, i.e. the latency. The latency is specific to the entity and is automatically set equal to the Aging parameter values of the SYS_DATA_DETAIL and WKLD_DATA_DETAIL tables.

Additional information

The latency setting limits the maximum age of data imported by an ETL task: if it is older than the aging policies of the SYS_DATA_DETAIL and WKLD_DATA_DETAIL tables it will be discarded by the warehousing engine and thus become useless.

For example, if the latency is set to six months, the warehousing process only takes into account the data present in a stage table whose timestamp is not older than six months. If the data loaded by an ETL task refers to metrics collected before this date, they will ben ignored by the warehousing engine and will not be reported.

These settings are necessary because the warehousing process has to deal with a great volume of summarizations; establishing limits prevents the process from becoming unmanageable.

To configure partition period for data tables

Another very important parameter that can be set in the Aging Configuration page is the partition period, which allows you to improve the performances of the  TrueSight Capacity Optimization DWH by using table partitioning. The partition period is expressed in days and defines the time span of each chunk in which the tables will be partitioned.

To tune this parameter you must consider the specific needs of your TrueSight Capacity Optimization deployment, because this setting speeds up the execution of SQL instructions by limiting it to one partition at a time. Partitions allow smaller indices to query, update and drop table rows. This feature is totally transparent to TrueSight Capacity Optimization users, and a good tuning of the parameter for each table can lead to major speed-ups.

For example, if you want to have a data retention period of three months but your users frequently perform analyses that are based on data from the last two weeks, you could set the partition period to 15 days. This way, most of the queries will be executed only on the last partition, which is much smaller than the whole table.

To add a new aging configuration

Click Add configuration to define a new aging configuration. The Add aging configuration form has the following fields:

Field

Description

Table name

Name of the database table for which you want to set an aging configuration

Column name

Name of the database column that contains the data timestamp

Aging

The number of days/weeks/months that should pass before the data is considered old and removed

Partitioned table

Select this option if you want to partition this table. Selecting this option has the following effect:

  • Enables the Partition period, Partition pattern, Forward, and Tablespace name options.
  • Disables the Additional conditions option.

Partition period

A drop-down menu that lets you select the time span for the partition chunk, e.g. year, semester, month, etc.

Tablespace name

The name of the tablespace to which the table belongs

Forward

How many days in advance the partition should be created

Additional conditions

Additional filtering conditions for data deletion that can be used only with unpartitioned tables. For example, in the case of stage tables aging, you can express the additional condition and status>0 to avoid deletion of data not yet processed by the warehousing engine.

Was this page helpful? Yes No Submitting... Thank you

Comments