This procedure allows you to re-index from scratch without causing downtime. The Front End uses existing indexes for search retrieval, and hotswapping replaces the old indexes with new ones.
To re-index the database without deleting the current indexes in memory:
- Log on to the FE server and enter the following commands to view the search_schema.xml file:se_confvi search_schema.xml
- Search for the <agent> tag to get the path to the <indexes> directory.In the following example, /exlibris/primo/p4_1/ng/jaguar/home/profile/indexes is the path of the indexes directory:<agents desc="collections of all search instances to be searched by jaguar" auto_deploy="active" swapping_type="slices"><agent port="9501" connection_type="local" host="primo1.corp.exlibrisgroup.com" active="true"><slices slice_count="1" name="main" active="true"></agent><slice></slices><path>/exlibris/primo/p4_1/ng/jaguar/home/profile/indexes/inst_1_index</path><load2ram>true</load2ram></slice></agents>
- On each SE server, move to the directory that contains the indexes directory and enter the following command to back it up, where <date> is the current date:cp -Rp indexes indexes.<date>
- Log on to the BE server as the primo user.
- Enter the following commands to delete the index files:be_bin./clean_indexes.sh
- From a Web browser, enter the URL for the Back Office server. For example:http://primo.exlibris.com:1601/primo_publishing/admin/acegilogin.jsp
- From the Back Office home page, click Primo Home > Monitor Primo Status > Process Monitoring.
- To re-index the database, click Execute for the Indexing_and_Didumean_and_Hotswapping process.
- Log on to the FE server as the primo user.
- Enter the following commands to monitor the job execution on the FE:fe_logtail -f library_server.logDo not restart the FE and SE servers so that you will be able to keep searching with the existing indexes.