Skip to content

searchisko/configuration

Repository files navigation

Searchisko configuration

Searchisko needs to be properly configured before it can be used. This project contains configuration of Searchisko running behind http://dcp.jboss.org/.

Installing configuration

Initialization steps:

  1. Get JBoss Developer artifacts from https://github.com/jboss-developer/www.jboss.org/tree/master/_dcp and copy them into data structure of configurations
  2. Create Elasticsearch index templates - index_templates/init_templates.sh
  3. Create Elasticsearch indexes - indexes/init_indexes.sh
  4. Create Elasticsearch mappings - mappings/init_mappings.sh
  5. Push all init data in the following order:
    • data/provider/init_providers.sh
    • data/config/init_config.sh
    • data/project/init_projects.sh
    • data/contributor/init-contributors.sh
    • data/query/init-queries.sh
  6. Initialize Elasticsearch rivers - rivers/init_rivers.sh

Note: each .sh script accepts commandline parameters which allows to configure location of Searchisko REST API or Elasticsearch search node http connector used by given script. Steps 2 - 5 can be performed by top level init_all_noriver.sh script.

Example - localhost:8080

  1. Optional: Clean environment (Searchisko data and embedded DB)

     rm -rf ~/.dcp_data/
     rm -rf $JBOSS_HOME/bin/searchisko.*
    
  2. Start Searchisko in local EAP. Read Development Docs.

  3. Push all init data

     cd configuration/
     ./init_all_noriver.sh
    

Example - OpenShift

  1. Push repo to Openshift

  2. Connect to Openshift Console

     ssh 5163d7b25973ca8ae4001fcf@dcp-jbossorgdev.rhcloud.com
    
  3. Stop EAP and the database and remove all Elasticsearch cluster data

     ctl_app stop
     rm -rf ~/app-root/data/search
    
  4. Delete data in the database

     # Recently, OpenShift does not seem to allow to stop only the app, but it stops the database as well,
     # we have to first start the app to make the database available.
    
     ctl_app start or do git push to openshift repo
    
     mysql dcp;
     delete from Config; delete from Contributor; delete from Project;
     delete from Provider; delete from Rating; delete from TaskStatusInfo;
     delete from Query;
     exit;
    
  5. Push all init data

     # Wait till app is up - visit main page of DCP
     cd ~/app-root/repo/configuration
     ./init_all_noriver.sh http://${OPENSHIFT_JBOSSEAP_IP}:8080
    
    • Optionally push other init data. For example if you have list of contributors or passwords for rivers. (Such init data is typically not shared in public repo.)
  6. Start Rivers

     cd ~/app-root/repo/configuration/rivers/
     ./init_rivers.sh
    

Versions management

As Searchisko is evolving and changing so does the configuration as well. It is important to pay attention to proper versions management.

  • Development: While new Searchisko version is in development phase, all related configuration updates are pushed into the master branch.

  • Release: As soon as new Searchisko version is released a new repository branch is created from master. The branch name must refer to Searchisko version.

  • Updates: Whenever there is made a change in configuration that apply to already released versions of Searchisko the update must be back-ported to all other relevant branches too.