Open topic with navigation
HPE IDOLconnectors aggregate content and metadata, process it and then index it into HPE IDOL Server in the form of fields. To improve HPE IDOL Server performance, divide these fields into the following groups:
Prevent from storing. Prevent HPE IDOL Server from storing fields that you do not want to query by setting the
CantHaveFieldsCSVs parameter in your HPE IDOL Server configuration file. Alternatively, add the
CantHaveFields parameter to your
DREADDDATA index action.
Index fields. Store fields that contain text that you want to query frequently as Index fields. HPE IDOL Server processes Index fields linguistically when it stores them. HPE IDOL Server applies stemming and stop word lists to the text, which allows HPE IDOL Server to process queries for these fields more quickly. Typically, you set up
DRECONTENT as Index fields.
Do not use Index fields to store URLs or content that you are unlikely to use. Also, when you query the value only in its entirety, it is more efficient to query with a field specifier (for example,
MATCH), than to store the data in an Index field.
HPE does not recommend indexing all fields in documents because it can potentially slow the indexing process, increase disk usage, and increase requirements.
Numeric fields. Store fields that contain numerical values or dates as numeric fields and numeric date fields. When HPE IDOL Server indexes these fields, it stores them in a fast look-up table in memory, which enables it to return the fields more quickly.
Field-check fields. If a large number of the documents that you want to store in HPE IDOL Server contain a field whose entire value is frequently used to restrict results, store this field as a
FieldCheckType field. When HPE IDOL Server indexes these fields, it stores them in a fast look-up table in memory, which enables it to return the fields more quickly.
Ordinary fields. By default HPE IDOL Server stores all fields that are not identified as special fields as ordinary fields.
You can query all stored fields using field specifiers in FieldText queries. You can also query Index fields using text queries.