Manticore Search 10.1.0: Prometheus Integration, Enhanced Monitoring, and Major Improvements

We're excited to announce the release of Manticore Search 10.1.0, featuring updates from May 2025, bringing built-in Prometheus integration, enhanced monitoring capabilities, and numerous stability improvements to make your search experience more reliable and efficient.

⚠️ Important Notice

Starting with version 10.1.0, CentOS 7 is no longer supported. We recommend upgrading to a supported operating system to maintain compatibility and security. Support for RHEL 10-based systems is currently in development.

🛠️ New Features and Improvements

This release includes several significant enhancements:

Built-in Prometheus Exporter (Issue #537): Added built-in support for exporting Prometheus metrics, making it easier to monitor and understand your search system. This feature was previously available only as a separate project: https://github.com/manticoresoftware/manticoresearch-prometheus.
KNN Index Management (Issue #3409): Added ALTER TABLE tbl REBUILD KNN command for easier migration of KNN indexes between versions.
Enhanced Cluster Recovery (Issue #1894): Improved cluster recovery with periodic seqno saving for faster node restart after crashes.

🔧 Critical Bug Fixes and Stability Improvements

This release addresses several important issues:

Search and Query Processing:
- Fixed wordform handling: user-defined forms now properly override auto-generated ones
- Fixed critical bug with fuzzy matching in specific cases (PR #538)
- Fixed inconsistent search results for delimiters in SQL vs JSON API (Issue #3091)
- Fixed incorrect sorting on json.field in facet search (Issue #3133)
- Fixed crash in tricky full-text query case (common-sub-term) (Issue #3356)
- Fixed crash when filtering aliased geodist() with JSON attributes (Issue #3080)
Performance and Resource Management:
- Improved performance by replacing DELETE FROM with TRUNCATE for distributed tables (Issue #2819)
- Enhanced auto disk chunk flush: skip flushing if optimization is running (Issue #3195)
- Fixed duplicate ID check for all disk chunks in RT tables (Issue #3313)
- Fixed crash with signal 11 during indexing (Issue #3306)
- Fixed crash related to remove_repeats() (Issue #3377)
- Fixed crash when creating a table with a KNN attribute but no model
API and Integration:
- Added support for latest Logstash and Beats (Issue #2400)
- Fixed space decoding in Buddy HTTP query parameters (Issue #3199)
- Added support for _random sorting in JSON API (Issue #3132)
- Fixed issues with uint64 document ID handling via JSON HTTP API (Issue #3382)
- Fixed incorrect handling of strings in HTTP JSON replies (Issue #3392)
- Fixed issue where non-existent @@variables always returned 0 (Issue #3109)
- Fixed incorrect results when filtering by id != value (Issue #3385)
Monitoring:
- Fixed small issues in SHOW VERSION output
- Fixed typo in disk chunk auto flush error message
- Added support for MCL 4.2.2; fixes errors with older storage formats (PR #3397)