Configure Relevance for Bloomreach Cloud
Introduction
Goal
Configure the Relevance Module for deployment in Bloomreach Cloud.
Background
Bloomreach Experience Manager's Relevance Module requires some specific configuration for deployment in Bloomreach Cloud.
Enable Elasticsearch in your Stack
If you plan to use Relevance features that require Elasticsearch*, please submit a request with Bloomreach Cloud support to enable Elasticsearch in your stack before your first deployment. Confirm the Elasticsearch version used in your stack and make sure your implementation project's Elasticsearch data store is properly configured for that version.
* Up to Bloomreach Experience Manager 13.4.1 and 14.0.0, using any Relevance feature requires Elasticsearch. As of Bloomreach Experience Manager 13.4.2 and 14.0.1, only the Experiments and Trends features require Elasticsearch.
Configure SQL Data Store maxAgeDays
Relevance relies on an SQL data store. When deploying your implementation project in Bloomreach Cloud, some specific configuration is required.
To keep backups small and optimize performance, configure the data store to keep records for a maximum of 30 days by setting the maxAgeDays property on the node /targeting:targeting/targeting:datastores/targeting:targetingdata:
/targeting:targeting/targeting:datastores/targeting:targetingdata: maxAgeDays: 30
Configure Client's Originating IP Address Resolution when using a CDN
If you have a CDN in front of your web application, the actual client's IP address must be retrieved from CDN-specific HTTP headers. Bloomreach Experience Manager's Relevance Module supports configuration of these headers through a context parameter in the site webapp's web.xml. You can specify multiple headers as a comma-separated string. If multiple headers are specified, the first available header will be used.
<context-param> <description>The remote ip address header</description> <param-name>http-forwarded-for-header</param-name> <param-value>X-Real-IP,X-Forwarded-For</param-value> </context-param>