
Photo by Igor Haritanovich: Pexels
ClickHouse May 2024 Release – v24.5
The open-source ClickHouse has a monthly release cycle includes new features, enhancements, and numerous bug fixes. On May 2024, ClickHouse May 2024 version 24.5 was released, and this version contains the following;
- 19 new features,
- 20 performance optimizations,
- 68 bug fixes.
For further details, please refer to the official ClickHouse docs here.
- v24.5 Source Code: GitHub Link
- v24.5 Release Webinar: Video
- Installation: ClickHouse Docs
This article will look at the important changes and features that are part of the ClickHouse 24.5 release.
New Features
Cross Joins
Added the possibility of doing CROSS JOIN in temporary files if the size exceeds limits (the blocks from the right table don’t fit into the memory) and compressing the correct table data while loading the data into RAM.
Map Data Type
Maps can now have Float, Array, Map, and Tuples as keys
S3
We can read data from archives located in local storage. Starting from this version, we can also read archives from S3 storage using the S3 table function.
This version has improved the performance of selection ({}) globs in S3.
Azure Blob Storage
Added support for azureBlobStorage function in ClickHouse to use Azure Workload identity to authenticate against Azure blob storage.
Performance Improvements
- Skip merging of newly created projection blocks while inserting the data.
- Add a new configuration called prefer_merge_sort_block_bytes to control the memory usage and speed up the sorting two times when merging the data parts. This is beneficial when there are many columns to merge.
- clickhouse-local is optimized to start faster
- Set index (Data skipping index) will be up to 1.5x faster
- Remove copying data when writing the data to the filesystem cache
- Native copy for Azure is available now
Bug Fixes
- Disabled enable_vertical_final setting by default due to an existing bug
- Fixed issues with backing up the table with multiple shards
- Fixed crash when calling some functions on const low-cardinality columns
- Fixed issues about queries with FINAL, which resulted in wrong results on tables that do not use adaptive granularity
- Fixed backup/restore operation of projection part in case projection was removed from table metadata, but part still has the projection
Conclusion
Overall, this release has many bug fixes that will enhance the stability of the ClickHouse database servers. Along with this, there are numerous performance enhancements and new features that could directly have a positive impact on end users. This feature has an extra focus on the Cloud storage-related enhancements/fixes and could continue to improve in these areas in the future.
References
https://clickhouse.com/docs/en/whats-new/changelog#-clickhouse-release-245-2024-05-30
https://clickhouse.com/blog/clickhouse-release-24-05
Using ClickHouse-Backup for Comprehensive ClickHouse® Backup and Restore Operations
You might also like:
- How to Monitor & Troubleshoot Log Contention in ClickHouse
- ClickHouse Indexing FAQs and Best Practices for High Performance
- ClickHouse Performance: Optimizing HASH GROUP BY and ORDER BY Queries
- How to optimize JOIN operations in ClickHouse?
- Building a Medallion Architecture with ClickHouse: A Complete Guide