cancel
Showing results for 
Search instead for 
Did you mean: 
angelborroy
Community Manager Community Manager
Community Manager

Alfresco Community Edition 26.2-preview introduces OpenSearch as a new search backend, a significant step forward from the long-standing Apache Solr 6 stack. With the release of alfresco-docker-installer 1.5.0, you can try this new configuration today.

What's New

ACS 26.2-preview ships with built-in support for OpenSearch 2 as an alternative to the traditional Alfresco Search Services (Solr 6). Instead of deploying a Solr-based search service, ACS connects to an OpenSearch server using the Elasticsearch subsystem. A dedicated batch-indexer service handles continuous document synchronization, polling the repository on a configurable schedule and keeping the index up to date.

Trying It with alfresco-docker-installer

The generator makes it straightforward to spin up a full stack with OpenSearch:

yo alfresco-docker-installer \
--acsVersion=26.2-preview \
--searchType=opensearch \
--ram=16 \
--serverName=localhost \
--port=80

This generates a Docker Compose environment with OpenSearch 2 and the alfresco-elasticsearch-batch-indexer, pre-configured and ready to run.

Important Caveats

This is early access, the ACS 26.2 release itself is a preview, and the OpenSearch integration is marked as experimental and not production-ready.

A few things to keep in mind before testing:

  • PostgreSQL is required. MariaDB is not supported with this configuration.
  • Memory footprint is higher. OpenSearch and the batch-indexer together require an additional 2 GB of RAM; a 16 GB system is recommended.
  • Single-node deployment only. This is not a clustered setup.
  • Security plugin is disabled. The OpenSearch security plugin is turned off for simplicity in this local/dev configuration.

Looking Ahead

OpenSearch represents Alfresco exploration of a modern, Elasticsearch-compatible search backend for both Community and Enterprise. The 26.2-preview cycle is the right time to evaluate it, report issues, and share feedback before it reaches a stable release.

Try it out and let us know how it goes!