Release 10.1A: OpenEdge Development:
Progress Dynamics Administration

How the Progress Dynamics framework uses the cache

Each object rendering procedure uses the class and entity information to render the objects in the application. This is how Dynamics checks availability of and fetches the information:

If the session property, StartupCacheClasses, is provided with the CVS list classes to cache for the session, Dynamics caches these classes in memory at startup.
The client checks if the class or entity information is available in memory, and uses it if found. Any of the previous calls in this client session might have fetched this information.
If the requested class or entity information does not exist in memory, the client checks its machine’s local disk for the class or entity information, and uses it if found. The session property, client_cache_directory (absolute path to an existing directory), can specify the search directory.
If no cache is generated or information is not found in the cache, the client makes a server request to fetch the information from the server. In an AppServer configuration this represents an extra AppServer request. In a slow network environment, the cost, in time, of a server round trip can be relatively expensive.
At session shutdown, Dynamics writes any cached entities out to disk.
Note: Dynamics writes only the entity cache out to disk, not the class cache.

If the client_cache_directory session property is specified, the cache is written to that directory. Otherwise, Dynamics writes the entity cache to the directory for the ry-clc product module. The session property, auto_dump_entity_cache, allows you to specify whether to have the shutdown process dump the entity cache to disk. For more information on these session properties, see the "Configuring the class and entity cache" section.