Search backend [VC 21.3 GEN]
Searching in Vidispine is implemented using either Solr or Elasticsearch as the backend. This allows functionality such as boolean operators, faceted searching, term highlighting, search term suggestions, etc.
Solr
Solr is the default search backend. It supports all features and works with both standalone Solr and SolrCloud.
Elasticsearch
Requirements
-
Elasticsearch 6.8.1
-
The analysis-icu plugin needs to be installed. If you are using Amazon Elasticsearch Service, this plugin is already installed.
-
Make sure to configure the search backend and that either the search backend URL or the
elasticsearchPathproperty is set properly. -
Run
vidispine elasticsearch initto initialize Elasticsearch. If an Elasticsearch instance is initialized already, the initialization will not create a new index, unless the--recreateargument is used when initializing. -
(Optional) Run
vidispine elasticsearch init --percolator-indexto initialize the “percolate query” index that is newly introduced in 5.6. This is useful for systems upgrading from an older version, and don’t want to re-initialize everything. (New in 5.6.) -
Running
vidispine elasticsearch checkcan be used to check if Elasticsearch is active already or should be initialized.
Limitations
The search interface is the same compared to using Solr (both syntactically and semantically), except that:
-
Collection-item join is not yet supported.
-
The
noescapeoption in the “version 1” search document is not supported. -
When searching using cursor for timed intervals, the same item or collection might be returned with different timespans on different result pages. This is because timespans are sorted independently, but grouped to items on a per-result-page basis.