ClickHouse continues to evolve with major improvements across data handling, query performance, security, and system management. One of the most significant updates is the production-ready JSON support, making ClickHouse a game-changer for semi-structured data analytics. ClickHouse now delivers super-fast analytics on JSON, outperforming Elasticsearch by 3x and MongoDB by at least 100x for analytical workloads. This makes it one of the most powerful solutions for handling structured and semi-structured data at scale.

Let us look at the important new features, changes and bug fixes in this release.

New Features and Enhancements

JSON and Dynamic Data Types Are Now Production-Ready

ClickHouse now fully supports JSON as a native data type, enabling high-performance querying of semi-structured data without complex transformations. Alongside this, Dynamic and Variant data types allow seamless data flexibility.

With this update, ClickHouse is now a serious competitor for analytical workloads traditionally handled by Elasticsearch and MongoDB. ClickHouse delivers:

• 3x faster query performance than Elasticsearch
• At least 100x faster performance than MongoDB for analytics on JSON data
• Columnar storage efficiency for better compression and lower storage costs
• Simplified querying with SQL rather than complex aggregation pipelines

For benchmarking details, see JSONBench.

Secure SSH Connections to ClickHouse

ClickHouse now supports SSH protocol for secure remote connections, making administration more reliable and secure.

Improved Table Function Handling with Cluster Alternatives

ClickHouse now automatically replaces table functions with their -Cluster equivalents when using parallel replicas, optimizing performance without requiring manual configuration.

Userspace Page Cache for Faster Data Access

A new Userspace Page Cache implementation allows ClickHouse to cache data directly in memory, reducing reliance on the OS page cache. This significantly improves query speeds in cloud environments and remote file systems.

Smarter CPU Scheduling for Fair Query Execution

A new concurrent_threads_scheduler setting improves CPU allocation between queries:

• round_robin (default) – Standard CPU allocation
• fair_round_robin – Ensures balanced distribution between INSERT and SELECT queries

This prevents smaller queries from being delayed when large workloads dominate resources.

New Functions for Advanced Analytics

ClickHouse introduces several powerful new functions:

• estimateCompressionRatio() – Analyze compression efficiency
• arraySymmetricDifference() – Identify differences across multiple arrays
• keccak256() – A cryptographic hash function commonly used in blockchain applications
• Iceberg-related functions for better table management

Performance Improvements

Query Condition Cache for Faster Repeated Queries

ClickHouse now temporarily caches query conditions, acting as in-memory indexes for repeated queries with similar filters. This reduces processing time and improves efficiency.

Optimized Data Handling and Memory Efficiency

Several internal optimizations reduce memory usage and boost performance:

• Smarter eviction of old cached data prevents memory bloat
• Improved handling of empty data parts avoids unnecessary file writes
• Faster Min/Max operations on Decimal32, Decimal64, and DateTime64
• Query compilation optimizations adapt execution to hardware architecture
• Improved memory management in remote reading buffers (e.g., S3, ClickHouse Cloud)

Faster ClickHouse Server Shutdown and Startup

ClickHouse has removed unnecessary delays in server shutdowns and optimized remote storage interactions, improving startup times in cloud environments.

Smarter Sorting and File Handling

• Optimized arraySort() improves array sorting speed
• Filesystem caching updates reduce unnecessary buffer allocations
• Updated Zstandard (zstd) compression improves compression and decompression speeds

System Enhancements and Security Updates

Improved Iceberg and DeltaLake Support

• Users can now specify Iceberg metadata files explicitly
• Atomic renames for TRUNCATE INTO OUTFILE improve data truncation efficiency
• AWS session token support simplifies authentication for DeltaLake tables

Security and Configuration Improvements

• Disallowed NaN/Inf as float settings to prevent invalid configurations
• Encrypted configuration support now extends beyond config.xml
• Replicated database credentials are no longer logged for security

Kafka and Cloud Storage Improvements

• Upgraded librdkafka to v2.8.0 for better Kafka performance
• Smarter S3 backup handling reduces unnecessary read/write operations

Conclusion

Faster, Safer, and More Reliable ClickHouse!

With these critical fixes, ClickHouse is now faster, more resilient, and even more powerful for real-world analytical workloads. From blazing-fast JSON queries to rock-solid distributed inserts, these updates ensure that your data is handled with precision, efficiency, and reliability.

If you’re looking for high-performance analytics without the headaches, now is the perfect time to upgrade!

Looking for Expert ClickHouse Solutions?

At Quantrail, we offer a fully managed ClickHouse service, seamless migration assistance, and dedicated service contracts to help businesses optimize their analytics stack. Whether you need a hassle-free ClickHouse deployment, expert support, or help transitioning from another database, we’ve got you covered. Let’s talk about how we can accelerate your analytics!

Contact

References

Official Changelog

Photo by zahid lilani: https://www.pexels.com/photo/golden-gate-bridge-san-francisco-2104742/