Lessons from our journey to enable global code search with Elasticsearch on GitLab.com

GitLab is working to enhance its search infrastructure by integrating Elasticsearch on GitLab.com to improve global search performance and code search availability for users, a feature already available to self-managed users. The primary challenges involve reducing the Elasticsearch index size, which currently occupies significant space, and enhancing administrative capabilities. Various strategies, such as adjusting NGram sizes and considering separate indexes, have been explored to optimize space usage and indexing speed. Despite setbacks like the inefficiency of forced merges and the need for full re-indexing when the schema changes, efforts continue to develop better administrative tools, enable selective indexing, and possibly allow disabling code indexing to provide more flexibility. Future plans include potentially sharding Elasticsearch per group to improve performance and resilience. GitLab remains committed to improving search capabilities while inviting community contributions and feedback.

No tracked trend matches for this post yet.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.