10.3. When to use the Cascade DataHub
Deciding whether or not to use the Cascade DataHub depends on a number of factors. In general the Cascade DataHub improves data handling and redistribution on the data server. Used with Cascade Connect, it can reduce network traffic, make domain independence possible, and notify a Windows client that a data source has died on the data server.
In addition to these considerations, although there are certain connection scenarios that can be programmed relatively easily without using the Cascade DataHub, others would be very difficult and time-consuming. The table below summarizes these scenarios. For more information on each scenario, please refer to the Making Multiple Connections chapter of this manual. Also see Windows client in read-only mode.
 | Any scenario beyond "simple" will require moderate to substantial programming effort, and the chances of making a mistake increase significantly. |
Table 10-1. Relative difficulty not using the Cascade DataHub
| Connection Mode | Number of data server computers | Number of Windows computers | Relative programming difficulty when not using Cascade DataHub |
|---|---|---|---|
| Unidirectional (read-only) | One | One | Simple. |
| Bidirectional (read/write) | One | One | Requires care. |
| Unidirectional (read-only) | One | Many (more than one) | Requires care. |
| Bidirectional (read/write) | One | Many (more than one) | Requires care. |
| Unidirectional (read-only) | Many (more than one) | One | Simple in QNX 4, otherwise difficult. |
| Bidirectional (read/write) | Many (more than one) | One | Difficult. |
| Unidirectional (read-only) | Many (more than one) | Many (more than one) | Difficult in QNX 4, otherwise very difficult |
| Bidirectional (read/write) | Many (more than one) | Many (more than one) | Very difficult. |