Indexing in Bedework

The indexing process

Bedework uses ElasticSearch as the indexer engine. As objects are created destroyed and updated they are indexed through calls to the indexer. The indexer can run as an embedded engine - as it arrives in the quickstart - or as an external service - which will be required for clustering.

ElasticSearch is completely unsecured and is not intended to be secured. For testing and debugging it is possible to enable the http interface but this may expose your index. DO NOT run in production with the ES http interface open. In embedded mode the only access should be through the api.

In non-embedded mode you can run ES on the same server or preferably a different server. ES - depending on the amount of data and types of search - can use a lot of memory for caching of filters. See the ES site for full instructions on running a non-embedded version.

Both embedded and non-embedded require a small number of changes to the configuration. These are already in place for the embedded version and are available in the quickstart in

bedework/config/elasticsearch/config/elasticsearch.yml

The changes made are

action.auto_create_index: false

Prevent ES from creating indexes implicitly. This can cause issues when reindexing.

Uncomment the following

node.max_local_storage_nodes: 1