System Options Caching page
IMPORTANT: If your organization or service provided has opted to use the Content Manager Administration client, this page will not be available in the Content Manager client System Options. To enable or disable any of these options, you will need to access the Content Manager Administration client. See the Content Manager Enterprise Studio - Dataset setup and maintenance - Content Manager Administration client help for further information.
Content Manager allows an Administrator to manipulate the amount of information cached into memory.
The business logic layer in the Content Manager architecture caches key items to improve performance. These caches reduce the number of round trips to the Content Manager Workgroup server, and also provide a business object buffer to collect update requests in memory until a commit point is reached (for example, when you click OK on a property sheet).
Content Manager is preconfigured with some "ideal" settings for these caches and mostly they do not need to be adjusted. In some circumstances, certain customers may see some performance benefits by making some adjustments to the standard settings. This page contains these settings.
Caution should be used when modifying these settings as under some circumstances changing values may have an adverse effect on system performance. Any changes to these settings should be thoroughly tested in a test environment before implementing in production.
In the first instance you can adjust the size of the caches for some of the key business objects in the Content Manager object model. There are two cache size settings for the object model caches: the size of caches when running a normal Content Manager client (e.g. TRIM.exe or TRIMDesktop.exe), and the size to make these caches when they are part of a Content Manager Service layer (e.g. the Content Manager Web Service) . Since a Web Service may be servicing more than one client, it is generally necessary to have larger caches for such applications. You can specify a size limit for records (how many separate record objects should be held at any one time by the Object Cache), locations, and then you can specify a default size that will be applied to all other object types. Think of the cache as holding a copy of each different record that you will be viewing or updating in a particular Content Manager session.
The numbers indicate the number of objects of a particular type that the object cache can hold.
These are the default values:
| Object | Client | Server |
|---|---|---|
| Records | 1024 | 8096 |
| Locations | 1024 | 2048 |
| Other Content Manager objects | 512 | 512 |
For a Windows Service application you can opt to allow the cache to continue to grow, or can configure the cache size specifically.
Fast access cache
The fast access cache is a performance option that only applies to the Content Manager Service API (or any third-party developed code that is designated as a web service). It is designed to improve performance in a scenario where there are many users doing large volumes of record downloads. It basically reduces contention on the Content Manager object cache which may happen if the Web Client object cache is filled up and needs to be constantly flushed when responding to read requests. Depending on the type of bottleneck encountered, one of these options may assist in improving performance. Select from:
- No Web Service Cache
- Cache all Commonly Used Objects
- Cache All Objects Except Records
- Cache All Objects
For the object types listed below, there is generally a low number of items and it is optimal to cache all instances of these objects in memory. To do this, when any of the objects are first required, all of the objects are downloaded from the Workgroup server. When all the possible objects are cached in the business layer, it means that searches can be performed against an in-memory list rather than requiring round trips to the Content Manager database. This is generally of great benefit to improving performance of the Content Manager business layer. Some customers do have unusual needs, however, and require a larger number of record types for example, than is suitable for this form of caching. In this case, when numbers get excessive and initial downloading is slow, it is better to turn off the "full caching" behaviour and have that object work in the normal fashion. The full cache is turned off for particular object types by checking the appropriate "Optimise performance for large numbers of" check box.
- Record Types - stops full caching of Record Types.
When you select this option, Content Manager will change the Record Types tab on the Search for Records dialogue box, which you can then use to filter your search results by Record Types.
Click Search - Find Records and click the tab to see its options.
See Record Search Record Types tab.
- Record Type Threshold - this sets the viewable threshold of Record Types listed in the Record Search Record Types tab.
- Thesaurus Terms - stops full caching of Thesaurus terms
- Lookup Items - stops full caching of Lookup sets
- Location Member Of Relationships - The Content Manager location object generally has a complete list of all the other locations that it shares a relationship with - traversing both up and down the location hierarchy. For larger organisations that have sections containing a large list of members (e.g. over 100), performance can be improved by turning off download of the downward hierarchy relationships. This optimisation also changes some elements of the user interface - preventing the display of the larger lists that are no longer cached.
IMPORTANT: If the Location Members of Relationships option is enabled the Has Members association for Location properties will be hidden.
In order to speed up the viewing of recently viewed or edited documents, a copy of these documents can be held in a temporary state on the computer that is running the business logic layer. If the cache is enabled, whenever a document is viewed, extracted or uploaded a copy is kept in the client document cache. You can specify a limit on the size of this cache - once it becomes full, older documents are removed to make room for more recently access documents. Once again, when running the business layer inside a Service component, a larger cache size may be warranted as it is servicing a larger number of users.
- When running in a Windows Client application:
- Use a document cache - select for Content Manager to cache documents on client machines. Select the size limit for the cache.
- When running in a Windows Service application:
- Use a document cache - select for Content Manager to cache documents on servers for web services. Set the size limit for the cache.