How historical data is retained in Data Warehouse

Data is maintained in Data Warehouse according to a schedule. As data gets older, the data record retention is reduced.

Data Warehouse retains historical data based on the data marts and granularity of the data, as shown in the following table.

Data mart Measured object Granularity Retention period
Performance marts Volumes and internal volumes Hourly 14 days
Performance marts Volumes and internal volumes Daily 13 months
Performance marts Application Hourly 13 months
Performance marts Host Hourly 13 months
Performance marts Switch performance for port Hourly 5 weeks
Performance marts Switch performance for host, storage, and tape Hourly 13 months
Performance marts Storage node Hourly 14 days
Performance marts Storage node Daily 13 months
Performance marts VM performance Hourly 14 days
Performance marts VM performance Daily 13 months
Performance marts Hypervisor performance Hourly 14 days
Performance marts Hypervisor performance Daily 13 months
Performance marts VMDK performance Hourly 14 days
Performance marts VMDK performance Daily 13 months
Performance marts Disk performance Hourly 14 days
Performance marts Disk performance Daily 13 months
Capacity marts All (except individual volumes) Daily 13 months
Capacity marts All (except individual volumes) Monthly representative 14 months and beyond
Inventory marts Individual volumes Current state 1 day (or until next ETL)

After 13 months (which is configurable), Data Warehouse retains only one record per month instead of one record per day for capacity, performance, and resource data in the following fact tables: