Manticore Search 17.5.1

Version 17.5.1

Released: February 7th 2026

Recommended Libraries

• Recommended MCL version: 10.2.0
• Recommended Buddy version: 3.41.0

If you follow the official installation guide, you don't need to worry about this.

❤️ We’d like to thank @pakud for their work on PR #4075.

Breaking Changes

• ⚠️ v17.0.0 Issue #4120 MCL 10.0.0: Added support for DROP CACHE. This updates the interface between the daemon and MCL. Older Manticore Search versions don't suport the newer MCL.
• ⚠️ v16.0.0 Issue #4019 Percolate query JSON responses now return hit _id and _score as numbers instead of strings, matching regular search; this is a breaking change for clients that relied on string type for these fields.

New Features and Improvements

• 🆕 v17.5.0 PR #130 MCL updated to 10.2.0: QWEN model support fixed and support for additional models added.
• 🆕 v17.3.0 PR #4186 Executor updated to 1.4.0, including an updated PHP version and the llm-php-ext extension.
• 🆕 v17.2.0 PR #4195 Updated MCL to 10.1.0; Added support for Qwen local embeddings models.
• 🆕 v17.1.0 Issue #3826 Jieba morphology instances are now shared across tables with the same configuration (mode, HMM flag, user dict path), greatly reducing memory use when many tables use Jieba (e.g. many empty tables no longer cause ~20 GB usage).
• 🆕 v17.0.7 Issue #2046 In RT mode, stopwords, wordforms, exceptions, and hitless_words can be set inline in CREATE TABLE (semicolon-separated values; wordforms/exceptions use > or => for pairs, with \ escaping), so tables can be created without external files; SHOW CREATE TABLE returns these inline values.
• 🆕 v16.3.0 KNN search now defaults to oversampling=3 and rescore=1, and supports omitting k so the query limit is used as the effective k; this reduces unnecessary oversampling and improves behavior when using SELECT * with KNN on columnar tables.
• 🆕 v16.2.0 PR #4088 Added searchd --quiet (-q) flag to suppress startup output (banner and precache messages), printing only errors; useful when starting and stopping searchd in a loop or from scripts.
• 🆕 v16.0.1 Issue #3336 HTTP connections are now persistent by default when using HTTP/1.1: clients no longer need to send the Keep-Alive header explicitly, reducing random connection failures in API clients (e.g. PHP, Go). To close a connection, the client sends Connection: close. HTTP/1.0 still requires Connection: keep-alive for persistence.

Bug Fixes

• 🪲 v17.5.1 Issue #3498 Fixed JOIN results returning empty or duplicated values when a column was both a string attribute and a stored field; the attribute value is now returned correctly.
• 🪲 v17.4.2 Issue #2559 Fixed joins on JSON string attributes (e.g. j.s) returning no matches; they now work like joins on plain string attributes.
• 🪲 v17.4.1 Fixed stored attributes not being evaluated in final when cutoff is set.
• 🪲 v17.2.10 Issue #425 Auto table creation (auto schema) now runs for REPLACE INTO as well as INSERT INTO, so tables are created on demand when missing.
• 🪲 v17.2.9 Issue #3226 Fixed incorrect GROUP BY results on columnar MVA with multiple GROUP BY columns by rejecting such queries with the same error as rowwise ("MVA values can't be used in multiple group-by").
• 🪲 v17.2.7 Issue #1737 Fixed highlight() with html_strip_mode=strip corrupting content by decoding entities and altering tags; original entity form is now preserved in the highlighted output.
• 🪲 v17.2.6 Issue #3203 Fixed ALTER TABLE REBUILD SECONDARY failing with failed to rename … .tmp.spjidx when the table had multiple disk chunks.
• 🪲 v17.2.5 Issue #3226 Fixed incorrect GROUP BY results on columnar MVA with multiple GROUP BY columns by rejecting such queries with the same error as rowwise ("MVA values can't be used in multiple group-by").
• 🪲 v17.2.4 Issue #4148 Fixed distributed queries returning stored fields from the wrong local index when agent tables contain duplicate document IDs; stored fields now follow index order and the first matching index is used for duplicates.
• 🪲 v17.2.3 Issue #4176 Fixed table rename breaking tables that use external stopwords, wordforms, or exceptions: ATTACH TABLE now migrates these files to the new chunk-based format and updates disk chunk headers, so after ALTER TABLE RENAME the daemon no longer reports missing external files on restart.
• 🪲 v17.2.2 Issue #1065 Added the expand_blended search option so that query tokenization applies the table's blend rules and expands blended variants (e.g. "well-being" → "well-being" | "wellbeing" | "well" "being"), allowing one query to match documents indexed with any of those forms.
• 🪲 v17.1.3 Issue #1618 Bumped Buddy to 3.40.7. Support for using Manticore with HikariCP and JPA/MyBatis (Spring Boot) is provided via Manticore Buddy, which implements the required MySQL wire protocol behavior.
• 🪲 v17.1.2 Issue #4128 Fixed MATCH with OR over the same phrase in different fields (e.g. (@name ="^New York$") | (@ascii_name ="^New York$")) returning matches from other fields; boolean simplification no longer drops field restrictions in that case.
• 🪲 v17.1.1 Issue #4131 Fixed ALTER TABLE with table-level settings (e.g. html_strip='1') failing on tables with auto-embeddings with "knn_dims can't be used together with model_name"; serialization now omits knn_dims when model_name is set.
• 🪲 v17.0.12 PR #4188 Fixed incorrect filter and filter tree evaluation in joined queries.
• 🪲 v17.0.11 Issue #3661 Fixed intermittent crashes when using columnar MVA64 attributes (e.g. during merge/optimize or when selecting stored/columnar attributes), by correcting docstore and blob-pool handling of columnar MVA64 data.
• 🪲 v17.0.10 Issue #3944 HIGHLIGHT and snippet generation now support REGEX in the query string, so REGEX matches are highlighted correctly when the same REGEX query is used in MATCH and HIGHLIGHT.
• 🪲 [v17....

Manticore Search 15.1.0

Version 15.1.0

Released: December 7th 2025

Recommended Libraries

• Recommended MCL version: 9.0.0
• Recommended Buddy version: 3.40.2

If you follow the official installation guide, you don't need to worry about this.

Breaking Changes

• ⚠️ v15.0.0 PR #4003 Updated MCL requirement to 9.0.0, bringing uncompressed float-vector storage, block-size change for KNN vectors, and non-buffered reads. This update changes the data format. Older MCL versions won’t be able to read it, but the new version can still read your existing columnar tables without any issues.

New Features and Improvements

• 🆕 v15.1.0 PR #3990 Improved disk chunk flushing log records by breaking the total time into clearer parts.
• 🆕 v14.7.0 Issue #3860 Updated Buddy requirement to 3.40.1, which includes an improvement in autocomplete: normalized bigram-separator characters to spaces and filtered out duplicate or malformed suggestion combinations to improve suggestion quality. It also includes a fix for invalid JSON errors in Kafka views and a fix for autocomplete so sorting combinations no longer throws errors when some keys are missing in the score map.
• 🆕 v14.6.0 Issue #615 Updated the Manticore Buddy requirement to version 3.39.1, which includes a fix for invalid JSON errors in the KNN plugin and allows Buddy handlers to override the HTTP Content-Type so /metrics now returns Prometheus text format (text/plain; version=0.0.4) instead of JSON, fixing scraping failures.
• 🆕 v14.4.0 PR #3942 Updated the Manticore Buddy requirement to 3.38.0, filtering out zero-doc suggestions, improving string-key handling with Ds\Map, changing memory-usage reporting format in Buddy from KB to bytes for greater precision and enhancing performance, stability, and maintainability.
• 🆕 v14.5.0 Issue #3329 Trim extra whitespace and new lines in JSON payloads when logging query requests — skip leading/trailing spaces to avoid logging malformed JSON.
• 🆕 v14.3.0 PR #3932 Improved handling of LOCK TABLES / UNLOCK TABLES: write locks now return warnings instead of errors, read locks are correctly shown in SHOW LOCKS, and the overall lock logic behaves consistently.
• 🆕 v14.2.0 Issue #3891 Added support for arbitrary filter expressions in JOIN ON clauses (not just equality comparisons), enabling queries like ... ON t1.id = t2.t1_id AND t2.value = 5.

Bug Fixes

• 🪲 v15.0.6 Issue #3601 Fixed a regression where the native Windows service failed to start when installed with a custom configuration path.
• 🪲 v15.0.5 Issue #3864 Fixed handling of "joined fields" in SQL-based sources so that an "end marker" gets set correctly when iterating joined hits.
• 🪲 v15.0.4 Issue #4004 Fixed a regression where HTTP /sql responses from the daemon incorrectly used the Content-Type: text/html header instead of application/json.
• 🪲 v15.0.3 Issue #2727 Fixed issue where grouping via GROUP BY / FACET on attributes generated with a JSON-to-attribute mapping failed to work.
• 🪲 v15.0.2 Issue #3962 Updated the Buddy requirement to 3.40.2, which adds fuzzy-search support to the /sql endpoint and includes other fuzzy-search fixes.
• 🪲 v15.0.1 PR #3922 Updated the documentation and tests related to supporting Logstash 9.2.
• 🪲 v14.7.6 PR #4002 Fixed KNN oversampling behavior: no longer compute rescored KNN distance when rescoring is not requested, and pass float-vector access hints through to the columnar storage layer.
• 🪲 v14.7.5 PR #3999 Fixed the “test_298” model to resolve a failing KNN-related test.
• 🪲 v14.7.4 Issue #3977 Tests on Windows sometimes produced .mdmp crash dumps — fixed so "ubertests" no longer leave minidumps after completion.
• 🪲 v14.7.3 Issue #3832 Fixed a bug where multi-query requests via /cli_json that include semicolons (e.g. combining SQL statements) failed — semicolons are no longer replaced with null characters before processing.
• 🪲 v14.7.2 Issue #1613 Documented the internal 32-bit mask used when counting ranking factors for Phrase/Proximity/NEAR operators and how it can undercount terms beyond the 31st keyword.
• 🪲 v14.7.1 PR #3992 Fixed missing error messages for HTTP UPDATE and DELETE requests on distributed tables handled via agents, so clients now correctly receive errors when operations fail.
• 🪲 v14.6.4 Issue #3478 Enhanced update validation to check that updated attributes do not conflict with full-text fields; updates now reject modifications if they target full-text fields.
• 🪲 v14.6.3 Issue #2352 Fixed an internal error when using distributed tables with persistent_connections_limit.
• 🪲 v14.6.2 Issue #3757 Fixed a bug where table-status counters misleadingly labeled with "_sec" (e.g. query_time_1min) actually reported milliseconds instead of seconds.
• 🪲 v14.6.1 Issue #3979 Fixed a bug where SHOW INDEX <name> STATUS search timing stats (search_stats_ms_*) did not match the values calculated from the query logs; the reported timings now reflect the real logged query times.
• 🪲 v14.5.8 Issue #3517 Updated HTTP header handling so that Buddy integrations can define or override headers instead of always forcing application/json.
• 🪲 v14.5.7 Fixed test 226 to enforce a stable result order, and corrected the GTest model for JSON mixed-array handling.
• 🪲 v14.5.6 Fixed Windows build by correcting a type mismatch in binlog.cpp, ensuring DoSaveMeta() now compiles correctly for Windows platforms.
• 🪲 v14.5.5 Issue #805 Issue #807 Issue #3924 Fixed inconsistent behavior in JSON-attribute faceting: aliasing a JSON array in a FACET clause now behaves the same as faceting on the array directly.
• 🪲 v14.5.4 [ Issue #3927](https://github.com/manticoresoftwa...

Manticore Search 14.1.0

Version 14.1.0

Released: November 7th 2025

❤️ We’d like to extend our heartfelt thanks to @ricardopintottrdata for their work on PR #3792 and PR #3828 — addressing issues around HAVING total counts and the filter with empty name error — and to @jdelStrother for their contribution with PR #3819, which improves handling of ParseCJKSegmentation when Jieba support isn’t available.

Your efforts help make the project stronger — many thanks!

Recommended libraries

• Recommended MCL version: 8.1.0
• Recommended Buddy version: 3.37.0

If you follow the official installation guide, you don't need to worry about this.

⚠️ Important

Version v14.0.0 updates the replication protocol. If you are running a replication cluster, you need to:

• First, cleanly stop all your nodes
• Then, start the node that was stopped last with --new-cluster, using the tool manticore_new_cluster in Linux.
• Read about restarting a cluster for more details.

New Features and Improvements

• 🆕 v14.1.0 Issue #3047 Added support for LOCK TABLES statements generated by mysqldump, improving the safety of logical backups.
• ⚠️ v14.0.0 PR #3896 Added a progress meter for donor and joiner nodes in replication SST shown in SHOW STATUS.
• 🆕 v13.16.0 PR #3894 Updated buddy from 3.36.1 to 3.37.0, which adds the "quorum" option.
• 🆕 v13.15.0 PR #3842 force_bigrams option to fuzzy and autocomplete plugins.

Bug Fixes

• 🪲 v14.0.1 Issue #3844 Fixed a crash caused by using max(ft field).
• 🪲 v13.15.13 PR #3828 Fixed an error when using an empty filter name.
• 🪲 v13.15.12 PR #3873 Updated buddy from 3.36.0 to 3.36.1 with an RT mode check in the EmulateElastic plugin.
• 🪲 v13.15.11 PR #3857 Added testing for Filebeat version 9.2.
• 🪲 v13.15.10 PR #3880 Tested automatic documentation translation after fixes.
• 🪲 v13.15.9 Issue #3783 Fixed an issue preventing native FreeBSD compilation.
• 🪲 v13.15.8 Fixed documentation translations.
• 🪲 v13.15.7 PR #3868 Updated executor from 1.3.5 to 1.3.6, adding support for the iconv extension.
• 🪲 v13.15.6 Fixed fuzzer build issue related to issue 3817.
• 🪲 v13.15.5 Issue #3644 Fixed a crash caused by specific full-text queries.
• 🪲 v13.15.4 Issue #3686 Fixed the issue where the full-text query "(abc|def)" didn't work as expected.
• 🪲 v13.15.3 Issue #3428 Added the ability to get the total number of results for queries using HAVING.
• 🪲 v13.15.2 Issue #3817 Added the option searchd.expansion_phrase_warning.
• 🪲 v13.15.1 PR #3848 Fixed replication transaction key generation and handling of conflicting transactions.
• 🪲 v13.14.0 Issue #3806 Fixed CALL SUGGEST not matching trigrams.
• 🪲 v13.13.8 PR #3839 Updated buddy from 3.35.4 to 3.35.5 to correct the regex for matching joins in SQL queries.
• 🪲 v13.13.7 Issue #3815 Updated buddy from 3.35.3 to 3.35.4 to fix an issue with negative IDs in REPLACE.
• 🪲 v13.13.6 PR #3830 Updated buddy from 3.35.1 to 3.35.3.
• 🪲 v13.13.5 PR #3823 Added a Grafana integration test.
• 🪲 v13.13.4 PR #3819 Fixed ParseCJKSegmentation when Jieba support is not available.
• 🪲 v13.13.3 PR #3808 Fixed error handling when filters are used in right-joined JSON queries.
• 🪲 v13.13.2 PR #3789 Validated KNN parameters.
• 🪲 v13.13.1 Issue #3800 Fixed build issues when compiling without cjk/jieba support.

Manticore Search 13.13.0

Version 13.13.0

Released: October 7th 2025

Recommended libraries

• Recommended MCL version: 8.1.0
• Recommended Buddy version: 3.35.1

If you follow the official installation guide, you don't need to worry about this.

⚠️ Important

Since the configuration file was updated, you might see a warning during the upgrade on Linux asking whether to keep your version or use the new one from the package. If you have a custom (non-default) configuration, it's recommended to keep your version and update the pid_file path to /run/manticore/searchd.pid. However, it should still work normally even if you don't change the path.

New Features and Improvements

• 🆕 v13.13.0 Added support for MCL 8.1.0 with SI block cache.
• 🆕 v13.12.0 Implemented the secondary_index_block_cache option, updated the secondary index API, and inlined sort accessors.

Bug Fixes

• 🪲 v13.11.8 Issue #3791 Fixed a race condition between checking and invoking a fired timer.
• 🪲 v13.11.7 Issue #1045 Fixed a systemctl warning on RHEL 8 during systemd updates by replacing the outdated path /var/run/manticore with the correct /run/manticore in the configuration. Since the configuration file was updated, you may see a warning during upgrade asking whether to keep your version or use the new one from the package. If you have a custom (non-default) configuration, it's recommended to keep your version and update the pid_file path to /run/manticore/searchd.pid.
• 🪲 v13.11.6 PR #3766 Added support for MCL version 8.0.6.
• 🪲 v13.11.5 PR #3767 Improved Chinese documentation translations and updated submodules.
• 🪲 v13.11.4 PR #3765 Fixed handling of aliased joined attributes.
• 🪲 v13.11.3 PR #3763 Fixed a crash that could occur during batched joins on string attributes and resolved an issue where filters sometimes didn't work with LEFT JOIN.
• 🪲 v13.11.2 Issue #3065 Fixed a crash when inserting data into a columnar table with index_field_lengths enabled.
• 🪲 v13.11.1 Issue #3751 Fixed a crash that occurred when deleting a document with embeddings enabled.

Manticore Search 13.11.1

Version 13.11.1

Released: September 13th 2025

Bug Fixes

• 🪲 v13.11.1 Issue #3751 Fixed a crash when deleting a document with embeddings enabled.

Manticore Search 13.11.0

Version 13.11.0

Released: September 13th 2025

The main highlight of this release is Auto Embeddings — a new feature that makes semantic search as easy as SQL.
No need for external services or complex pipelines: just insert text and search with natural language.

What Auto Embeddings Offer

• Automatic embedding generation directly from your text
• Natural language queries that understand meaning, not just keywords
• Support for multiple models (OpenAI, Hugging Face, Voyage, Jina)
• Seamless integration with SQL and JSON APIs

Recommended libraries

• Recommended MCL version: 8.0.3
• Recommended Buddy version: 3.35.1

If you follow the official installation guide, you don’t need to worry about this.

New Features and Improvements

• 🆕 v13.11.0 PR #3746 Added "query" support in JSON queries for embeddings generation.
• 🆕 v13.10.0 Issue #3709 The manticore-server RPM package no longer owns /run.
• 🆕 v13.9.0 PR #3716 Added support for boolean_simplify in the configuration.
• 🆕 v13.8.0 Issue #3253 Installed sysctl config to increase vm.max_map_count for large datasets.
• 🆕 v13.7.0 PR #3681 Added support for alter table <table> modify column <column> api_key=<key>.

Bug Fixes

• 🪲 v13.10.5 PR #3737 The scroll option now correctly returns all documents with large 64-bit IDs.
• 🪲 v13.10.4 PR #3736 Fixed a crash when using KNN with a filter tree.
• 🪲 v13.10.3 Issue #3520 The /sql endpoint no longer allows the SHOW VERSION command.
• 🪲 v13.10.2 PR #3637 Updated the Windows installer script.
• 🪲 v13.10.1 Fixed local time zone detection on Linux.
• 🪲 v13.9.2 PR #3726 Duplicate IDs in columnar mode now correctly return an error.
• 🪲 v13.9.1 PR #3333 The manual is now auto-translated.
• 🪲 v13.8.6 PR #3715 Fixed an embeddings generation issue when all docs in a batch were skipped.
• 🪲 v13.8.5 PR #3711 Added Jina and Voyage embedding models, along with other automatic embeddings-related changes.
• 🪲 v13.8.4 PR #3710 Fixed a crash on joined queries with multiple facets.
• 🪲 v13.8.3 Fixed an issue where delete/update commits in a transaction with multiple statements at the _bulk endpoint were not counted as errors.
• 🪲 v13.8.2 PR #3705 Fixed a crash when joining by non-columnar string attributes and improved error reporting.
• 🪲 v13.8.1 PR #3704 Fixed a crash in query embeddings when no model was specified; added embeddings string to master-agent communication; added tests.
• 🪲 v13.7.2 PR #Buddy#589 Removed the default IDF hack for fuzzy search.
• 🪲 v13.7.1 Issue #3454 Fixed incorrect scroll decoding with large 64-bit IDs.
• 🪲 v13.6.9 Issue #3674 Fixed JDBC+MySQL driver/connection pool issue with the transaction_read_only setting.
• 🪲 v13.6.8 PR #3676 Fixed a crash on an empty result set returned by the embeddings model.

Manticore Search 13.6.7

Version 13.6.7

Released: August 8th 2025

Recommended MCL version: 8.0.1
Recommended Buddy version: 3.34.2

If you follow the official installation guide, you don’t need to worry about this.

New Features and Improvements

• 🆕 v13.6.0 Issue #2226 Support for explicit '|' (OR) in PHRASE, PROXIMITY, and QUORUM operators.
• 🆕 v13.5.0 PR #3591 Automatic embedding generation in queries (Work in Progress, not production-ready yet).
• 🆕 v13.4.0 PR #3585 Fixed logic to prefer buddy thread count from buddy_path config if set, instead of using the daemon value.
• 🆕 v13.3.0 PR #3577 Support for joins with local distributed tables.
• 🆕 #3647 Added support for Debian 13 "Trixie"

Bug Fixes

• 🪲 v13.6.7 Issue #3524 Fixed an issue with saving generated embeddings in row-wise storage.
• 🪲 v13.6.6 Issue #3563 Fixed issues with Sequel Ace and other integrations failing due to "unknown sysvar" errors.
• 🪲 v13.6.5 Issue #3467 Fixed issues with DBeaver and other integrations failing due to "unknown sysvar" errors.
• 🪲 v13.6.4 Issue #3524 Fixed an issue with concatenating multi-field embeddings; also fixed embedding generation from queries.
• 🪲 v13.6.3 Issue #3641 Fixed a bug in 13.6.0 where a phrase lost all bracketed keywords except the first one.
• 🪲 v13.6.2 Fixed a memleak in transform_phrase.
• 🪲 v13.6.1 Fixed a memory leak in version 13.6.0.
• 🪲 v13.5.2 Issue #3651 Fixed more issues related to full-text search fuzz testing.
• 🪲 v13.5.1 Issue #3560 Fixed a case where OPTIMIZE TABLE could hang indefinitely with KNN data.
• 🪲 v13.4.2 Issue #2544 Fixed an issue where adding a float_vector column could corrupt indexes.
• 🪲 v13.4.1 Issue # Added fuzz testing for full-text parsing and fixed several issues found during testing.
• 🪲 v13.3.1 Issue #3583 Fixed a crash when using complex boolean filters with highlighting.
• 🪲 v13.2.7 Issue #3481 Fixed a crash when HTTP update, distributed table, and wrong replication cluster are used together.
• 🪲 v13.2.6 PR #3567 Updated manticore-backup dependency to version 1.9.6.
• 🪲 v13.2.5 PR #3565 Fixed CI setup to improve Docker image compatibility.
• 🪲 v13.2.4 Fixed handling of long tokens. Some special tokens (like regex patterns) could create words that are too long, so they are now shortened before use.

Manticore Search 13.2.3

Version 13.2.3

Released: July 8th 2025

Blogpost: https://manticoresearch.com/blog/manticore-search-13-2-3/

This version focuses on making Vector Search faster, more memory-efficient, and better suited for large-scale data, while also improving compatibility and stability.

Breaking Changes

• ⚠️ v13.0.0 Updated the KNN library API to support empty float_vector values. This update doesn’t change the data format but does increase the Manticore Search / MCL API version.
• ⚠️ v12.0.0 PR #3516 Fixed a bug with incorrect source and destination row IDs during KNN index training and building. This update doesn’t change the data format but increases the Manticore Search / MCL API version.
• ⚠️ v11.0.0 Added support for new KNN vector search features like quantization, rescoring, and oversampling. This version changes the KNN data format and the KNN_DIST() SQL syntax. The new version can read older data, but older versions won't be able to read the new format.

New Features and Improvements

• 🆕 v13.2.0 PR #3549 Fixed an issue with @@collation_database that caused incompatibility with certain mysqldump versions
• 🆕 v13.1.0 Issue #3489 Fixed a bug in Fuzzy Search that prevented parsing of certain SQL queries
• 🆕 v12.1.0 Issue #3426 Added support for RHEL 10 operating systems
• 🆕 v11.1.0 Added support for empty float vectors in KNN search
• 🆕 v10.2.0 Issue #3252 log_level now also adjusts Buddy's logging verbosity

Bug Fixes

• 🪲 v13.2.3 PR #3556 Fixed "oversampling" option parsing in JSON queries
• 🪲 v13.2.2 Fixed crash logging on Linux and FreeBSD by removing usage of Boost stacktrace
• 🪲 v13.2.1 Issue #3298 Fixed crash logging when running inside a container
• 🪲 v12.0.2 Improved accuracy of per-table statistics by counting in microseconds
• 🪲 v12.0.1 PR #3522 Fixed a crash when faceting by MVA in a joined query
• 🪲 v11.0.3 Issue #3502 Fixed a crash related to Vector Search Quantization
• 🪲 v11.0.2 Issue #3493 Changed SHOW THREADS to display CPU utilization as an integer
• 🪲 v11.0.1 Fixed paths for columnar and secondary libraries
• 🪲 v10.2.7 Added support for MCL 5.0.5, including a fix for the embeddings library filename
• 🪲 v10.2.6 Issue #3469 Applied another fix related to issue #3469
• 🪲 v10.2.4 Issue #3469 Fixed an issue where HTTP requests with bool queries returned empty results
• 🪲 v10.2.5 Changed default filename for the embeddings library and added checks for the 'from' field in KNN vectors
• 🪲 v10.2.3 PR #3464 Updated Buddy to version 3.30.2, which includes PR #565 about memory usage and error logging
• 🪲 v10.2.2 Issue #3470 Fixed JSON string filters, null filters, and ordering issues in JOIN queries
• 🪲 v10.2.1 Fixed a bug in dist/test_kit_docker_build.sh that caused the Buddy commit to be missing
• 🪲 v10.1.4 Fixed a crash when grouping by MVA in a joined query
• 🪲 v10.1.3 Issue #3434 Fixed a bug with filtering empty strings
• 🪲 v10.1.2 PR #3452 Updated Buddy to version 3.29.7, which resolves Buddy #507 - error on multi-query requests with fuzzy search and Buddy #561 - Fix for metrics rate, required for Helm Release 10.1.0
• 🪲 v10.1.1 Updated Buddy to version 3.29.4, resolving #3388 - "Undefined array key 'Field'" and Buddy #547 - layouts='ru' may not work

Manticore Search 10.1.0

Version 10.1.0

Released: June 9th 2025

Blogpost: https://manticoresearch.com/blog/manticore-search-10-1-0/

This version represents an update with new features, one breaking change and numerous stability improvements and bug fixes. The changes focus on enhancing monitoring capabilities, improving search functionality, and fixing various critical issues that affected system stability and performance.

Starting with version 10.1.0, CentOS 7 is no longer supported. Users are encouraged to upgrade to a supported operating system.

Breaking Changes

• ⚠️ v10.0.0 Issue #540 BREAKING CHANGE: set layouts='' as the default for fuzzy search

New Features and Improvements

• 🆕 v10.1.0 Issue #537 Added built-in Prometheus exporter
• 🆕 v9.8.0 Issue #3409 Added ALTER TABLE tbl REBUILD KNN
• 🆕 v9.7.0 Issue #1778 Added automatic embeddings generation (we're not officially announcing it yet, as the code is in the main branch but still needs more testing)
• 🆕 v9.6.0 Bumped KNN API version for auto-embeddings support
• 🆕 v9.5.0 Issue #1894 Improved cluster recovery: save seqno periodically for faster node restart after crash
• 🆕 v9.4.0 Issue #2400 Added support for latest Logstash and Beats

Bug Fixes

• 🪲 v10.0.1 Fixed wordform handling: user-defined forms now override auto-generated ones; added test cases to test 22
• 🪲 v9.8.2 Fix: updated deps.txt again to include packaging fixes in MCL related to the embeddings library
• 🪲 v9.8.1 Fix: updated deps.txt with packaging fixes for MCL and the embeddings library
• 🪲 v9.7.3 Issue #3306 Fixed crash with signal 11 during indexing
• 🪲 v9.7.2 Issue #3109 Fixed issue where non-existent @@variables always returned 0
• 🪲 v9.7.1 Issue #3377 Fixed crash related to remove_repeats()
• 🪲 v9.6.3 PR #3411 Fix: use dynamic versions detection for telemetry metrics
• 🪲 v9.6.2 Fix: small fix in SHOW VERSION output
• 🪲 v9.6.1 Fix: crash when creating a table with a KNN attribute but no model
• 🪲 v9.5.16 Issue #3342 Fixed issue where SELECT ... FUZZY=0 didn't always disable fuzzy search
• 🪲 v9.5.15 PR #3397 Added support for MCL 4.2.2; fixes errors with older storage formats
• 🪲 v9.5.14 Issue #3392 Fixed incorrect handling of strings in HTTP JSON replies
• 🪲 v9.5.13 Issue #3356 Fixed crash in tricky full-text query case (common-sub-term)
• 🪲 v9.5.12 Fixed typo in disk chunk auto flush error message
• 🪲 v9.5.11 Issue #3195 Improved auto disk chunk flush: skip saving if optimization is running
• 🪲 v9.5.10 Issue #3313 Fixed duplicate ID check for all disk chunks in RT table using indextool
• 🪲 v9.5.9 Issue #3132 Added support for _random sorting in JSON API
• 🪲 v9.5.8 Issue #3382 Fixed issue: couldn't use uint64 document ID via JSON HTTP API
• 🪲 v9.5.7 Issue #3385 Fixed incorrect results when filtering by id != value
• 🪲 v9.5.6 PR #538 Fixed a critical bug with fuzzy matching in some cases
• 🪲 v9.5.5 Issue #3199 Fixed space decoding in Buddy HTTP query parameters (e.g., %20 and +)
• 🪲 v9.5.4 Issue #3133 Fixed incorrect sorting on json.field in facet search
• 🪲 v9.5.3 Issue #3091 Fixed inconsistent search results for delimiters in SQL vs JSON API
• 🪲 v9.5.2 Issue #2819 Improved performance: replaced DELETE FROM with TRUNCATE for distributed tables
• 🪲 v9.5.1 Issue #3080 Fixed crash when filtering aliased geodist() with JSON attributes

Manticore Search 9.3.2

Version 9.3.2

Released: May 2nd 2025

This release includes multiple bug fixes and stability improvements, better table usage tracking, and enhancements to memory and resource management.

❤️ Special thanks to @cho-m for fixing build compatibility with Boost 1.88.0 and @benwills for improving documentation around stored_only_fields.

• 🪲 v9.3.2 Fixed an issue where the "Show Threads" column showed CPU activity as a float instead of a string; also fixed a PyMySQL result set parsing error caused by a wrong data type.

• 🪲 v9.3.1 Issue #3343 Fixed leftover tmp.spidx files when the optimize process was interrupted.

• 🆕 v9.3.0 PR #3337 Added a per-table command counter and detailed table usage statistics.

• 🪲 v9.2.39 Issue #3236 Fix: prevent table damage by removing complex chunk updates. Using wait functions inside the serial worker broke serial processing, which could damage tables.
  Reimplemented autoflush. Removed external polling queue to avoid unnecessary table locks. Added "small table" condition: if the number of documents is below the 'small table limit' (8192) and no Secondary Index (SI) is used, flushing is skipped.

• 🪲 v9.2.38 Fix: skip creating Secondary Index (SI) for filters using ALL/ANY on string lists, without affecting JSON attributes.

• 🪲 v9.2.37 Issue #2898 Added backtick support for system tables.

• 🪲 v9.2.36 Fix: use a placeholder for cluster operations in legacy code. In the parser, we now clearly separate fields for table and cluster names.

• 🪲 v9.2.35 Fix: crash when unquoting a single '.

• 🪲 v9.2.34 Issue #3090 Fix: handling of large document IDs (could previously fail to find them).

• 🪲 v9.2.33 Fix: use unsigned integers for bit vector sizes.

• 🪲 v9.2.32 Fix: reduce peak memory usage during merge. docid-to-rowid lookups now use 12 bytes per document instead of 16 bytes. Example: 24 GB RAM for 2 billion docs instead of 36 GB.

• 🪲 v9.2.31 Issue #3238 Fix: incorrect COUNT(*) value in large real-time tables.

• 🪲 v9.2.30 Fix: undefined behavior when zeroing string attributes.

• 🪲 v9.2.29 Minor fix: improved a warning text.

• 🪲 v9.2.28 Issue #3290 Improvement: enhanced indextool --buildidf

• 🪲 v9.2.27 Issue #3032 With the Kafka integration, you can now create a source for a specific Kafka partition.

• 🪲 v9.2.26 Issue #3301 Fix: ORDER BY and WHERE on id could cause OOM (Out Of Memory) errors.

• 🪲 v9.2.25 Issue #3171 Fix: crash caused by segmentation fault when using grouper with multiple JSON attributes on an RT table that has multiple disk chunks

• 🪲 v9.2.24 Issue #3246 Fix: WHERE string ANY(...) queries failed after RAM chunk flush.

• 🪲 v9.2.23 PR #518 Minor auto-sharding syntax improvements.

• 🪲 v9.2.22 Issue #2763 Fix: global idf file was not loaded when using ALTER TABLE.

• 🪲 v9.2.21 Fix: global idf files can be large. We now release tables sooner to avoid holding unnecessary resources.

• 🪲 v9.2.20 PR #3277 Improvement: better validatation of sharding options.

• 🪲 v9.2.19 PR #3275 Fix: build compatibility with Boost 1.88.0.

• 🪲 v9.2.18 Issue #3228 Fix: crash during distributed table creation (invalid pointer issue).

• 🪲 v9.2.17 PR #3272 Fix: multiline fuzzy FACET issue.

• 🪲 v9.2.16 Issue #3063 Fix: bug in distance calculation when using the GEODIST function.

• 🪲 v9.2.15 Issue #3027 Minor improvement: support for Elastic's query_string filter format.