Data aggregation
Storage, Fabric, and Virtualization are just a few of the subsystems that NetBackup IT Analytics supports in terms of performance metric gathering.
When regulating the growth of performance metrics data, NetBackup IT Analytics purges old data periodically using customizable retention parameters. But if the retention periods need to be longer than the defaults, this data can eventually take up a lot of space on the disc where the database is stored.
Additionally, when the data in the underlying tables expand exponentially, data retrieval becomes difficult.
With the release of version 11.2.02 for Capacity and Switches sub-systems, NetBackup IT Analytics now provides Data Aggregation for performance metrics data to manage the aforementioned scenarios. In the subsequent releases, it will also be expanded to include Virtualization and Cloud subsystems.
Higher data retention durations are made possible by data aggregation, which aggregates data for performance measures without increasing the database's required disc space.
The existing data in the table is aggregated as per the preselected aggregation metrics (average, max, min, count, sum, standard deviation etc.) for a specific time resulting in reduced number of records.
For instance, NetBackup IT Analytics allows the collection of performance statistics for logical units within 30 seconds. This can lead to 28 million records each day for 10,000 LUNs. Without aggregation, this table's retention duration of one year will produce 10 billion records. While for the same retention, this data can be reduced by ten times with aggregation enabled.
The following are the advantages of data aggregation.
Disk space: Depending on the frequency of data collection, data aggregation can potentially shrink the quantity of raw data by up to 10 times, thus lowering the impact of the enormous performance metrics data.
Quick report response: Report queries against the database processes more quickly since aggregated values have already been persisted.
Data retrieval: Data retrieval is made simpler by lowering the size of the data.