Reference the Architecture page for a diagram of the OntoPortal system components.
The search index is maintained continuously in Solr. When an ontology is submitted, the ontology is re-indexed, and the new index entries replace the existing index entries for that ontology.
Re-indexing of individual ontologies can be initiated via OntoPortal web Admin interface and it can also be done from the linux command line interface on the Appliance. If the whole index becomes corrupted or if solr schema is changed then all the ontologies must be re-indexed.
Solr runs on port 8983 which is blocked by local firewall because it not inteded to be accessed directly. Most of the typical solr administrative operation can be accomplished from the appliance console using solr API. If you need to acces solr web admin interface from outside of the applince then you would need to set ssh port fowarding.
There are 4 index cores:
sudo su - ontoportal cd /srv/ontoportal/ncbo_cron bin/ncbo_ontology_index -a -l logs/reindexing_all.log
Re-indexing of all ontologies process deletes all data from the index and starts populating it from scratch. During this time Appliance search operations will not work as expected because of the incomplete data until it is fully populated. If you have a large ammount of ontologies and unable to take site down for maitanence then you have an option to re-index all ontologies using alternate core and swap cores after after re-indexing is complete.
The first step is to re-index all ontologies using secondary core:
sudo su - ontoportal cd /srv/ontoportal/ncbo_cron bin/ncbo_ontology_index -a -l logs/reindexing_all.log -c http://localhost:8983/solr/term_search_core2
then swap cores after reindexing process is complete:
You can re-index specific ontologies you specify.
bin/ncbo_ontology_index -o STY,SNOMED -l logs/reindexing_STY_SNOMED.log