Symptom

PSRM does occur performance degradation, when opening a folder or part that does contain too many documents. The reason for that being is that parsing the XML document, that does make a part simply takes too long, if a part does contain too many documents. Firstly the XML iterator takes long and secondly, every document is being instantiated to check for authorizations, status and availability of the physical document.

(Recommendation from the author: The check for availability should actually be switched off – which is a simple task in redefining one method of the document backend class only!)

Solution

When saving a part, the current number of documents should be evaluated against a maximum threshold. If reached the current part should be closed and a new one created.

Requirement

Within a PSRM folder many parts can be stored, of which the latest is the current part and open for insertion of documents. To avoid degradation of performance, this current part should be closed and a new part opened, once a certain number of documents are reached.

The implementation should support the different options of access to the part, when declaring documents online and in background. Scenarios could be:

• Online

1. From within the current part, create a document and declare it into the very location, then saving the folder
2. From within PSRM create a document and declare it using the “location finder”
3. When dragging and dropping documents from the organizer into the part, then saving the folder

• Offline

1. From programs that create a document, then use the API’s to declare a document into a part, locating it through opening the target folder. Sources could be either single upload or scan, or mass-upload by migration programs

As a feature, the number of documents should be customizable on the Element Type. An extra Connection Parameter will be added and utilized in the evaluation method determining if a part should be closed or not.

Please notice the download links for details:

• The documentation
• The sources as a NUGG file