This article provides some details about how you can plan for and estimate your database requirements for Mashup Center.
General database configuration
When you configure database persistence for Mashup Center, we recommend that you configure two separate databases for Lotus Mashups and MashupHub. Configuring two separate databases ensures higher performance as well as added flexibility and tuning options. However, if you must configure a single database for both components, be sure that they use separate schemas for a cleaner, more support-friendly environment.
Estimating the size of your databases
When estimating the size of your databases, the first thing you should consider is the expected user load. Keep in mind that the majority of content that actually gets stored in your database is user-driven content, for example content that users actually create and store in the database. Certain content from static pages is stored in your database as well, but typically the storage requirements for static pages is light. Static pages generally require a lot of database reads, but that content is well cached and should not be a significant factor in your initial load estimates. Another factor to consider is future growth and the addition of new features after you move into a production environment.
To help you with your estimates, you should understand which components require higher amounts of storage space. For MashupHub, those components are typically widgets and data sources that you store in the catalog. Widgets and data sources can grow quite large, sometimes as large as 1MB or more, for a single widget or Microsoft Excel spreadsheet. For Lotus Mashups, pages require the most storage space. Although a single page takes up only about 10KB of file space, the space required for pages can grow exponentially if you have a large number of users who are creating multiple pages. For example, if 10,000 users each create five pages, your data storage requirements can reach as much as 500MB, not including any required overhead.
Although the size of your database depends on all the factors described above, as a general rule, we estimate that you should plan for an average of 1 to 2GB for each database in a production environment.