Hyland’s Alfresco Search Enterprise module integrates Elasticsearch, enhancing search capabilities with improved performance, flexible indexing and AI readiness, while offering cost-effective, scalable deployment options.
Hyland’s new Alfresco Search Enterprise module uses Elasticsearch as the underlying search engine. This shift replaces the previous search functionality, Search Service’s Solr, and provides some distinct advantages for Hyland’s Alfresco customers.
Let’s explore how Elasticsearch and Alfresco now work together for a more powerful solution.
Alfresco Search Enterprise uses the standard Elasticsearch product, which has several advantages for deployment:
Alfresco Search Enterprise supports the free versions of both Elasticsearch and OpenSearch. This means no additional licenses are needed to take advantage of the new capabilities.
Alfresco Search Enterprise does not require any modifications to Elasticsearch. This allows for a wide range of deployment options:
Additionally, with standardized software, organizations can enjoy simpler maintenance and expanded access to expertise, which helps reduce complexity of shard maintenance.
Because standard software is used, Hyland customers can more easily find expertise to help them get set up and managing their indexes. There is no Alfresco-specific knowledge required.
Because standard software is used, Hyland customers can choose to engage with many different providers of managed cloud services for Elasticsearch or OpenSearch on many different cloud platforms, such as Elasticsearch itself and AWS.
Alfresco Search Enterprise’s “live indexer” that indexes new and changed content is built as an out-of-process, event-driven extension. An additional “bulk indexer” component is available to take care of heavy indexing workloads, indexing directly from the Alfresco database for super fast indexing speeds. This means indexing is now fully separate from Hyland’s Alfresco Content Services, allowing customers to easily scale up their indexing capacity in times of a large influx of content, without having to provision a whole Alfresco Content Services instance for this. Additionally, indexing volume no longer takes capacity from an Alfresco Content Services instance.
A great architecture with flexible scaling and many different deployment options is useless if the software does not deliver the other capabilities an organization needs. Here are some of the functional advantages Alfresco Search Enterprise offers:
With support for most of the query operations of CMIS and AFTS, Alfresco Search Enterprise offers a drop-in replacement for Search Services without having to rewrite queries. Alfresco takes care of translating queries into Elasticsearch queries.
With Search Services, content the current user did not have access to was filtered out of the result set after the query was run. This didn’t serve users well because not only did it slow down a search action, but it also made the number of returned documents unpredictable. Users had to gauge whether they found all relevant documents or ran up to the maximum result limit of Solr.
With Alfresco Search Enterprise, content access is indexed along with the content itself. This allows Alfresco Content Services to add access rights to the query itself. This means no postprocessing filtering is necessary, and the number of documents returned is not altered by access rights filtering for the current user. This also results in a small but measurable improvement in search query performance.
Using the bulk indexing component, metadata indexing speeds are 10x faster than with previous search solutions. A recent test showed that indexing of 2 billion documents with complex and variable metadata could be achieved in 25 hours. This indexing performance can be achieved by deploying parallel indexing components when a significant bulk indexing is required.
While Alfresco Search Enterprise, like Search Services, still operates on an “eventual consistency” model, our performance tests show that indexing new and changed content is orders of magnitude faster regardless. With Alfresco Search Enterprise, content can be found much faster than with Search Services.
In our performance tests, queries against Alfresco Search Enterprise outperform queries against Search Services and significantly outperform the database-based TMDQ queries. This enhances the experience for the end user, and it also means it is possible with Alfresco Search Enterprise to have larger content repositories with the same query performance as Search Services.
Elasticsearch is designed to work with very large datasets, removing the search engine as the bottleneck for Alfresco scalability. With the distributed Live Indexing components, the repository is no longer a bottleneck for indexing performance. Alfresco Search Enterprise allows for better performance with larger repositories.
When training AI systems, the quality of the data is of utmost importance. Alfresco Search Enterprise makes it possible to prepare your data for embedding your content into AI solutions like large language models (LLM).
Make the most of content and the power of AI with:
Alfresco’s flexible metadata models make it easy to add the right metadata to your content. These metadata fields are indexed in Elasticsearch and available for any processing in or after Elasticsearch.
While Alfresco Content Services generates the right field mappings in Elasticsearch based on the data models, it is possible to add or change mappings and have the processing done on content and content metadata. This allows customers to tailor the data stored in Elasticsearch for their specific AI needs.
According to Elasticsearch, its product is the world’s most downloaded vector database. Alfresco Search Enterprise indexes content and metadata in Elasticsearch, making the data in the content repository available for any solution using Elasticsearch’s vector capabilities.
Customers can use the Alfresco index in Elasticsearch to power any kind of embedding from an LLM to Elasticsearch, such as the one Elasticsearch contributed to LangChain.
Hyland’s Alfresco Search Enterprise module revolutionizes search functionality with enhanced performance and flexibility. Discover how this powerful solution can transform your content management. For more details and personalized support, contact Hyland to explore how we can meet your unique needs.
