Improving rate table performance

For each rate table, you can choose whether to load the table into memory or access it from the database. By default, PolicyCenter loads rate tables into memory. Loading the rate table in memory can provide quick access. Accessing the table on disk is slower, but loading a large table into memory may slow system performance even more.

To decide whether or not to load the table into memory, consider how large the table will be and how often it will be referenced. Consider database lookup if:

The table has multiple thousands of rows and is referenced infrequently. For example, referenced only on a small percentage of quotes.
The table is very large (multiple tens of thousands of rows) and is only referenced once or twice per quote.

For anything smaller, or for a table referenced numerous times to quote a policy, use memory lookup, if possible. Database lookup can result in significant slowdown.

Rating management consults the lookup setting when promoting the rate book status from Draft to Stage. Therefore, you can only change the lookup setting when the rate book is in Draft status.

A rate table can be included in multiple rate books but can have only one setting for lookup. You can only change the lookup setting in the first rate book that included the rate table, if it is in Draft status.

Performance depends on the rate table content, not just the row count. For a table that is small enough to fit in memory, in-memory will usually outperform a call out to the database. For database-queried tables, if the query tends to hit the correct row every time, database performance is better than if the query frequently goes through the relaxing process. A query hitting on the first try usually corresponds to a table which is mostly filled out, with very few null values. On the other hand, you can expect relaxation to occur often if the table has many null values.