Notes on Database Disk Sizing
Factors Affecting Disk Sizing Estimates
Currently there is no detailed information on how to scale the expected size of the BI data warehouse from a source ERP system. The reason for this is that the size of the BI data warehouse depends on various factors:
• The number of modules implemented
• The size of the source ERP tables transformed
• The number of ERP sources
• The volume of daily and historical ERP data
• The degree of customization of the BI modules
Because these factors are difficult to predict beforehand, we currently give guidelines on expected data warehouse size based on existing customer data.
Estimation of Initial Database Size
If necessary, refer to
Reference: Estimation of Initial Database Size for an approximate method of estimating the initial BI database size. This estimate is approximate in nature, and should be used only as a means of getting an order-of-magnitude estimate on database size.
Allowance for Database Growth
Depending on the daily transaction volumes, the number of BI modules, and the customizations, the BI database grows on a daily basis; this depends on the daily scheduled load jobs. Make an allowance up-front for this in terms of disk capacity.
The recommendation is to have a disk capacity of 2-3 times the estimated size of the BI database. The rule of thumb is that if the initial BI DW is based on converting three years of historical data and you assume that the same rate of transactions are passing through the source ERP database, then expect the BI DW itself to grow by a similar volume to the initial size of the DW over the coming three years.
Also, daily or weekly monitoring of the BI database size provides ongoing measurement of growth rates, so provide adequate storage to manage this growth.
Allowance for Database Backup
Ensure a recommendation for a tape-drive backup or other appropriate backup media.