At line 1 changed one line |
The search configuration allows you to customize the database used for indexed searches. You can use this to adjust usernames that are used to find the VFS areas to index, as well as configuring your own custom database to hold the search criteria. CrushFTP will periodically keep the search DB records updated, as well as tracking renames, deletes, and uploads directly in the search DB. The list of usernames are empty by default, and the search DB does not operate then. |
__IMPORTANT: due to the security updates since CrushFTP version 10.5.2+ any JDBC driver jar file needs to be placed into the CrushFTP10/plugins/lib/ directory, or it won't load.__ |
\\ |
\\ |
The search configuration allows you to customize the database used for indexed searches. You can use this to adjust usernames that are used to find the VFS areas to index, as well as configuring your own custom database to hold the search criteria. CrushFTP will periodically keep the search DB records updated, as well as tracking renames, deletes, and uploads directly in the search DB. The list of usernames are empty by default, and the search DB does not operate then. The usernames are used to scan their VFS to see what they can see, and only those folders are indexed and added to the DB. The indexing interval runs with the parameter search_index_interval in minutes. All the other field should be left as is with their defaults. |
At line 7 added 24 lines |
|
If you want to use an alternate DB then you will need to configure these fields:\\ |
\\ |
• search_db_driver: specify the driver you want to use like... "com.mysql.jdbc.Driver"\\ |
• search_db_driver_file: specify the path to where the JDBC driver jar file is at like... "./mysql.jar"\\ |
• search_db_url: specify the URL for the connection to the DB like... "jdbc:mysql://127.0.0.1:3306/crushftp_search"\\ |
|
|
|
Content based search indexing for various document file types is also possible, it will require an external free tool, Tika. The |
only drawback is the indexing process is slow and CPU intensive. Will require use of external SQL database. |
|
|
* first download tika, rename it to tika-app.jar, copy into CrushFTP installation directory root |
|
http://www.apache.org/dyn/closer.cgi/tika/tika-app-1.10.jar |
|
* configure 'Search Config' to use an SQL database, do a connection test, create tables with the SQL query strings on config panel |
|
http://www.crushftp.com/crush7wiki/Wiki.jsp?page=MySQL |
|
* set search_index_interval to a fairly short time interval ( 1 minute) |
|
* specify in search_index_usernames field the Crush accounts of which VFS folders you wish to be indexed |