ObjectCacheSizeLimit
Integer (prefix multiplier)
80M
Purpose
The ObjectCacheSizeLimit parameter specifies the maximum amount of memory that is allocated by the JADE client node for caching persistent objects in the client machine.
The optimum value for this parameter varies, depending on the:
-
Amount of physical memory available on the client machine
-
Average amount of data modified in a typical database transaction
-
Size of the working set of persistent objects that is accessed frequently
The minimum value is 3M bytes and the maximum value is two‑thirds of the physical memory on the client machine (with a maximum of 4G bytes on 32‑bit JADE systems).
When JADE detects a parameter value that is outside the allowed range, the value is changed to be the appropriate minimum or maximum, a message is output to the jommsg.log file, and the JADE initialization file is updated. Node initialization then continues normally.
Parameter is read when …
The application server or the standard (fat) client node is next initialized.
Hints
When the JADE client reads an object, it is loaded from the server node into the client node object cache. When that object is read again, it is obtained directly from the cache. However, lock and write operations still need to coordinate the status of local object buffers with the server node so that the client node is guaranteed to lock or modify the latest object buffer.
Having a large number of objects in the client's cache could significantly reduce the network traffic in certain multiuser applications, thus improving the overall performance of the client.
The impact of this parameter is less for single user applications, where loading buffers into the client node cache does not involve communications across workstations. Additionally, if the working set of the application is relatively big or if it changes frequently, the effect of the cache operation is less noticeable.
Use the persistentCacheStats parameter in the logObjectCaches method of the
Increasing the value of the ObjectCacheSizeLimit parameter may reduce cache misses. (For details about sampling JADE nodes to obtain statistical information, see