| title | authors | |
|---|---|---|
Solr Server (Daemon) |
|
This Debian package and Ubuntu package is a preconfigurated Apache Solr server running as a daemon providing important settings like the integration of the thesaurus editor and ontologies manager, settings for better performance, disabled logging and security settings and a more current Solr version than the packages of the Debian or Ubuntu standard repositories.
Disabled logfiles: we don't want to write each search query to Solr logs. If you want to switch on logging for debugging purposes, switch on file and console in the config file /var/solr/log4j.properties
Automatic commits to the index after 15 seconds after adding or update of documents (autocommit=15000)
Automatic start on booting running as a daemon in Debian GNU/Linux or Ubuntu Linux.
Automatic memory settings: In most cases no manual setting of Java virtual machine options is needed anymore. Allows the Java VM to use as much RAM as possible on this server, so you won't have problems because of default Java Virtual Machine (JVM) maximal RAM settings (option -Xmx) if indexing large amounts of data or large documents.
Disabled swappiness, so the system will only swap if necessary. So it doesn't do so to optimize RAM for running software swapping the Solr index and search caches automatically after some time because they are not used for some time. Why? Even if some parts of the Solr index and caches in RAM are not used for a long time (i.e. if search isn't used for the night or some days) and that RAM could be used by other software meanwhile to read hundrets of MB or some GB from Swap on slower harddisks to RAM again because of using again while the first search after long time would lead to timeouts and errors on maybe important searches, which then could take tens of seconds longer.
For security reasons access to the Solr search server is only possible from the same computer. So access is only possible from localhost, so that if you set a password to the User Interfaces module solr-php-ui and the search apps nobody without an account on your computer or an account for a service on your computer can read all the data from Solr.
To enable Solr remote admin access from other computers than localhost you have to edit jetty-http.xml and delete the default="127.0.0.1" from the config option "host". Then restart Solr by service solr restart.
Warning: You don't want to enable access to an unprotected Solr server with the possibility to read, add, change or delete all indexed data for everybody on the intranet or internet! So if the computers are part of a network you can not fully trust, you have to protect the IP of the Solr server or the Solr port for example by a firewall.
There are additional fields and stemming configured in the Solr schema. You can read the XML schema config in /var/solr/data/core1/conf/managed-schema which is based on the Solr example config set /opt/solr/server/solr/configsets, so you can use a diff tool to compare and see the config additions.
Additionally the ETL and search tool adds & uses some additional fields which are created automatically using the Solr dynamic fields feature configured for the schema because of type endings like _b _s or _tt. You can see such additional fields using the table view.