NB : The feature is temporarily unavailable for has been re-enabled since Datafari 4.0. It will be integrated again in a later release of Datafari.2.
| Info |
|---|
To see the feature described bellow in the search view of Datafari, you must be connected to Datafari with a user having either the "SearchExpert" role, or the "SearchAdministrator" role. Refer to the Types of users for more details. |
Once you are connected as a "SearchExpert" or a "SearchAdministrator", if you perform a search in Datafari, you will notice two news new icons at the right of each result:
These icons are used to elevate, drop or remove an elevate boost of the associated document for the current query:
: Is used to elevate the associated document. It will add a boost to the document for the current query, that will make it appear as the first result of the listin the firsts results.
When you click on this button, Datafari modifies the elevate.xml file located in [Datafari_Home]/solr/solr_home/FileShare/conf to either add or move the document as to the first of the list and then restart the Solr core. This is the reason why you will have a delay before the search view is reloaded and the document appears as the first of the results.list of the elevated docs for this query if it is not already in, or to raise it one step up.
If you elevate several documents with this method, notice that it each added document will always be the last elevated doc which will be on top of the results list. It works like a stack where the last added element is added on top of the stack so it has the priority on the previous inserted one which also has the priority on the previous etc etc.
This feature is intended to quickly and easily elevate only one or very few documents. As on each click the Solr core is reloaded and that it may be annoying to find the doc you want to elevate or to wait after each click to elevate several docs, you will prefer to manage the elevate configuration through the Boosting documents from the admin UI.
: Is used to remove the elevate boost of the associated document for the current query. If the document has an entry for the current query in the elevate.xml file located in [Datafari_Home]/solr/solr_home/FileShare/conf, it will be removed from the list of elevated documents and the Solr core reloaded.
As for the elevate feature, this search view incorporated feature is intended to quickly and easily remove elevate boost on only one or very few documents. For better control and ergonomic interface, you will prefer the Boosting documents from the admin UI. inserted at the end of the list.- : Is used to drop an elevated document one step down in the list.
: Is used to remove an elevated document from the list.
: Indicates that the corresponding document is already in the list of elevated documents for the current query. The displayed number indicates the position of the document in the list, it is refreshed in real time when the user add, raise, down or remove documents from the elevated docs.
To avoid ressource consumption and interruption of service, every doc elevation that a user performs thank to this UI will only be applied to the elevate.xml file. But the file will not be automatically uploaded to the zookeeper nodes and the Solr core will not be realoaded. It means that the elevations will not take effect until the configuration is manually uploaded to the zookeeper nodes and the Solr core is reloaded !
To do this, two options are available:
- Use the documents boosting admin UI to select the query on which documents have been boosted and click on the confirm button. It will automatically upload the elevate.xml file to the Zookeeper nodes and reloaded the Solr core. But be careful ! the process will disable any searches on Datafari for few secondes for and index with less than 100k documents to several minutes for an index of more than 1M docs. The more documents the index contains the more the process time will be long.
- Manually upload the Solr configuration to the Zookeeper nodes and reload the Solr core thanks to the Zookeeper admin UI. Like with the first option be careful ! the process will disable any searches on Datafari for few secondes for and index with less than 100k documents to several minutes for an index of more than 1M docs. The more documents the index contains the more the process time will be long.