Releasing SIREn 1.2: Major Performance Boosts for Nested Queries

Releasing SIREn 1.2: Major Performance Boosts for Nested Queries

We are delighted to announce the release of SIREn 1.2. You can get the latest release from the download page.

SIREn is a schemaless structured document search system that combines free text search with structured search over arbitrary JSON data. The system is an extension of Lucene/Solr, the world’s leading search engine technology.

The focus of this release is on performance. We have created a new indexing mode which can be up to 70 times faster for certain types of nested queries.

How did we achieve this? After evaluating common use cases, we recognized that most of them do not require querying over attribute names. Based on this, we created the hybrid indexing mode which trades off support for this capability with a performance boost. We call this mode hybrid since it combines the field based indexing approach of Lucene with the node based indexing approach of SIREn. While the original mode will still be available, the hybrid mode will be the default indexing mode.

Besides this, v1.2 includes enhancements such as support for custom datatypes in nested faceting, an upgrade to the latest Lucene/Solr 4.8.1 as well as some optimizations and bug fixes. You could see the changes log in the distribution package for more details.

We are also hard at work on the upcoming release (v1.3). This release will introduce something that we have been receiving a lot of requests for – a SIREn plugin for Elasticsearch. We will be demonstrating how nested querying in SIREn is significantly more memory efficient relative to standard support in ElasticSearch.

You might want to sign up to our SIREn User Group to find out more about all of this, be informed about updates and to participate in discussions about structured document search.

Renaud Delbru

No Thanks / Already Signed Up