Best Practices for Usage Data Maintenance
By default, usage data maintenance is not enabled. If your reporting system has newly imported files that were never aggregated and you are unsure how many months of data you may need for reporting, it is recommended that you leave the Usage Data Maintenance settings disabled at first.
After aggregation is finished and you have analyzed your Flexera Analytics (Cognos) reports, you can specify the appropriate hourly, daily, and monthly Usage Data Maintenance settings. These three settings can have different values based on business needs—for example, your enterprise may need to retain 12 months of monthly usage data, six months of daily usage data, and three months of hourly usage data in the system.
While enabling Usage Data Maintenance, be sure to consider aggregation scheduling. When you save the Usage Data Maintenance settings, the data cleanup service will be invoked immediately. While the usage data cleanup service is running, if a scheduled or manually triggered aggregation starts, the aggregation may try to save the usage data and the data cleanup service may try to delete the usage data for the same timeframe. This could result in undesired behavior. Before running aggregation, check the reporting flexnet.log to determine whether the data cleanup service is in process.
The following example demonstrates how the aggregation process works:
| 1. | FlexNet Manager for Engineering Applications builds hourly aggregation for seven days (default configuration). The input is transactional data. |
| 2. | FlexNet Manager for Engineering Applications build daily aggregation for the same seven days. The input is hourly aggregation. |
| 3. | FlexNet Manager for Engineering Applications build monthly aggregation in a similar way, taking input from daily aggregation. |
In this scenario, if you re-run aggregation for a past period of time—for example, six months from now—and you have a three-month retention policy for hourly aggregation, it may result in creating hourly aggregation for the time beyond the retention date just to support the re-aggregation for daily and/or monthly aggregation. Normally the “extra” data will be deleted later, but the best practice is to disable the usage data maintenance settings prior to aggregation that may overlap in time with the retention intervals.
It is recommended you define your data-retention time intervals such a way that there is a “buffer time interval” that ensures that the usage data for cleanup would not overlap with any possible usage data for reports or aggregations.
Usage Data Cleanup Service Log Messages
When running each type of data cleanup service (hourly, daily, or monthly), the following line appears in the log to indicate that the data cleanup service is running:
Starting Hourly Cleaning activity for <from-date> to <to-date>
When the data cleanup service is finished, it will log the following:
Stopping Usage Data Cleanup Service.
If the data cleanup service found no usage data, it logs the following:
Hourly cleanup activity found no data to clean up.
See Also