How To
Documentation
    Learn More

      Fusion 5.1.3 Release Notes

      Release date: 18 June 2020

      Component versions:

      Solr 8.4.1

      ZooKeeper 3.5.6

      Spark 2.4.5

      New Features

      Fusion

      • Fusion 5.1.3 adds the Connector Plugins Repository API. This API is used to install and update connector plugins from the built-in plugin repository. For example, the following command describes a successful installation of a connector plugin with the API:

        curl -u admin:Password123 -X POST "http://fusion-cluster:6764/connectors/plugins?id=lucidworks.fs"
        {
          "pluginId" : "lucidworks.fs",
          "status" : {
            "action" : "INSTALL",
            "status" : "SUCCESS",
            "message" : "Plugin lucidworks.fs was installed successfully."
          }
        }

      Improvements

      Fusion

      • Support for service:// and extservice:// is added to the REST Query Index Stage. This is consistent with the Rest Call Jobs, which supports calls to other Kubernetes services. For example, service://<serviceName>/<path> or extservice://<serviceName>/<path>.

      • The Solr Subquery query pipeline stage now supports grouped results from Solr grouped queries. Access the groupValue and the documents for a group in the Javascript stage:

        groupValue = ctx.get("subquery-results").grouped.get("Category_s").getGroups().get(0).getGroupValue()
        docs = ctx.get("subquery-results").grouped.get("Category_s").getGroups().get(0).getDoclist().docs

        Now, Solr Subquery’s SearchResponse has a new field, grouped, which is a map of group field names to GroupCollections.

      Other changes

      Fusion

      • The Push connector is removed. As an alternative, directly access the index endpoints.

      • Java-based services have moved to distroless Docker images. As a result, debugging tools must be run remotely.

      Bug fixes

      Fusion

      • IMPORTANT: Fixed a bug that prevented the FTP connector from working.

      • Fixed a bug that caused long data values in job statuses to exceed the allowed character limit.