Symptom

There is sometimes a requirement to have to create a new Solr index (or a re-index) for Curator Server. This is can be seen when there are errors within Process Engine workflows or some of the user interfaces where a mismatch when mapping Solr assets to the database is reported. The cause of this can be many things: a common cause is where asset deletion takes place in the database but the delete isn't pushed into Solr. 

Other reasons might include an update applied to the database directly, which isn't advisable unless IPV is first consulted and the change approved. This would also require the creation of a new Solr index.

Resolution

The time it takes to complete the index depends on the number of records within the database. Another contributing factor is the amount of memory available on the server. The bigger the database, the longer it will take to complete. 

Records are made up of assets, folders and metadata. If the database has millions of records, the process will take hours.

In order to create a new index, the steps are:

• On the server where Curator Server is installed, browse to http://localhost:8983/
• On the left-hand menu, select Core Selector and then the latest/newest curator core (for example curator_2_2) The activity core is not to be selected in this case.
  
• From the drop-down menu, select Dataimport.
  
• From within the /dataimport options, make sure the following are ticked: Clean, Commit, Auto-Refresh Status. Then click Execute.
  
• You'll see a live update on the top right of the same screen as the index compiles. Eventually, you'll be met with the green text Indexing completed - this message will contain some information on how many documents have been processed.
  
• Now Curator is ready to be used again.

If for any reason the indexing doesn't complete, please raise a ticket with Technical Operations and we'll be happy to assist.