Manticore Search 7.4.6: Enhanced Data Visualization with Kibana Integration, Performance Optimization, and Key Fixes

We are excited to announce the release of Manticore Search 7.4.6, which features powerful data visualization capabilities through its integration with Kibana, significant performance optimizations for join operations and histograms, as well as numerous fixes that enhance stability and reliability.

The standout feature of this release is the long-anticipated integration with Kibana (#832), which greatly simplifies and accelerates the data visualization process. You can now create stunning dashboards, graphs, and visualizations directly based on Manticore Search data, making data analysis even more accessible.

This integration unlocks new opportunities for log monitoring and analysis based on Manticore Search.

Try it out for yourself: we have prepared a demo showcasing the capabilities of the Manticore-Kibana combination, allowing you to compare performance against Elasticsearch. In tests with a dataset of approximately 40 million documents, the same dashboard loads in 7 seconds with Manticore Search compared to 18 seconds with Elasticsearch!

Additionally, here are some key improvements in this update:

JOIN query optimization (#2995): Data merging during complex JOIN operations has been enhanced, significantly speeding up execution times.

Histogram optimization (#3039): The EstimateValues method for histograms now operates more quickly, improving analytics on large datasets.

Columnar storage enhancements (#77):

Optimized data block reuse when creating filters with multiple values.

Min/max values have been added to attribute metadata.

Pre-filtering of filter values based on min/max has been implemented, speeding up query execution.

Consistency across platforms (#1727): Discrepancies in floating-point calculation accuracy between arm64 and x86_64 architectures have been resolved, ensuring uniform results on all platforms.

This release addresses several critical issues:

Fixed expression handling in JOIN queries when using attributes from both tables (Commit 73ac).

Corrected incorrect results when using avg() in JOIN queries (#2915).

Resolved errors in results caused by implicit truncation during JOIN grouping (#2996).

Fixed a memory leak in the JOIN query cache (#3042).

Corrected request parameter processing in JOIN queries involving JSON (#3052).

Addressed a crash when shutting down the service during active disk chunk merging (#3031).

Fixed a daemon crash caused by a full-text query with a single NOT operator and a custom ranker (#3126).

Eliminated a potential vulnerability in the CJSON library (#3128) ❤️ Special thanks to @tabudz for the PR.

Resolved an issue where IN(...) could yield incorrect results (#3037).

Addressed issues with the ATTACH TABLE command (#3054).

Fixed duplicate records after bulk replacement of IDs close to the maximum value of 2^64 (#3088).

Resolved a bug where setting max_iops / max_iosize in version 7.0.0 could degrade indexing performance (#3038).

Fixed an issue where setting diskchunk_flush_write_timeout=-1 at the table level did not function correctly (#3087).

In this release, we have improved versioning. Starting from Manticore 7.4.6, we are adopting full semantic versioning. This will make version numbers easier to understand and predict for developers.

Key benefits include: clearer understanding of changes, tracking development progress, and independent versioning of packages.

For more details, see our article on versioning in Manticore.

Boost 1.87.0 support (#3099): Support has been added for the latest version of the Boost library. ❤️ Thanks to @cho-m for the PR.

Consistency in error messages (#3079): Inconsistencies in error messages have been corrected for improved developer experience.

We welcome your feedback!

Join the conversation on our forum.

Report bugs and propose improvements on GitHub.

Connect with us on Slack.

Reach out to us at contact@manticoresearch.com.

Special thanks to @cho-m, @tabudz, and everyone who contributed to this release! ❤️