Requirement to expose larger number of properties as columns
Hello, We would like to get some clarity on the following.
With relation to saving business case data (That doesn’t not belong to a separate logical entity) that would be used for searching & reporting following is our understanding.
1. It is not a good practice to expose larger number of scalar properties as columns on the worktable. Consensus is to keep this to a minimum like less than 15 columns.
2. Then the alternate approach would be to create a data type or a declare index.
3. In conversations on this topic, we have heard following concerns about using a declare index for this requirement (Requirement to expose larger number of properties as columns).
a. Reservations on using declare indexes due to experience of earlier versions of Pega b. Some concerns about creating multiple declare indexes having an impact on performance. What is the cap (Number of declare indexes) on using declare indexes.
4. We are not keen on creating a data type to store business case details due to following reasons
a. Creating a data type is not the right approach since data can be relatively easily accessed and modified intentionally or unintentionally in production. Ideally in an enterprise application if the data needs to be saved in an entity of their own, this DB entity tables needs to be properly encapsulated with proper privileges. Even when a data type row is unintentionally touched, the updated user details will be changed.
Hello, We would like to get some clarity on the following.
With relation to saving business case data (That doesn’t not belong to a separate logical entity) that would be used for searching & reporting following is our understanding.
1. It is not a good practice to expose larger number of scalar properties as columns on the worktable. Consensus is to keep this to a minimum like less than 15 columns.
2. Then the alternate approach would be to create a data type or a declare index.
3. In conversations on this topic, we have heard following concerns about using a declare index for this requirement (Requirement to expose larger number of properties as columns).
a. Reservations on using declare indexes due to experience of earlier versions of Pega b. Some concerns about creating multiple declare indexes having an impact on performance. What is the cap (Number of declare indexes) on using declare indexes.
4. We are not keen on creating a data type to store business case details due to following reasons
a. Creating a data type is not the right approach since data can be relatively easily accessed and modified intentionally or unintentionally in production. Ideally in an enterprise application if the data needs to be saved in an entity of their own, this DB entity tables needs to be properly encapsulated with proper privileges. Even when a data type row is unintentionally touched, the updated user details will be changed.
b. Having a data type(s) to store 1. larger number of scalar properties & 2. list properties introduce overhead while performing CRUD operations
c. In rare situations data can go out of sync with BLOB & data types
d. Data types does not have significant reporting performance benefits compared to declare indexes since both approaches contain data in dedicated tables.
In summary, we are keen on using declare indexes in following scenarios 1. Create a declare index to expose larger number of properties 2. Create declare indexes to expose list values
There are not enough support material and defined solution on this topic and hence the request.
Thank you!