On October 13, TmaxTibero officially released Tibero PatchSet04, encompassing updates for Tibero 6 (6.7.4) and Tibero 7 (7.2.4).
This patchset delivers major advancements in parallel processing, data compression, cluster reliability, and data-handling efficiency, further reinforcing Tibero’s position as a mission-critical enterprise database.

Additionally, this release integrates several preliminary technological foundations for the forthcoming AI Database “Tibero 8” (codename), positioning it as a cornerstone for next-generation intelligent database systems.

New Feature of PatchSet 04

1. TCC (Tibero Columnar Compression) — Enhanced Columnar Data Compression

The newly introduced TCC feature surpasses the limitations of traditional row-oriented compression, enabling columnar compression optimized for analytical and large-scale data workloads.
Data is grouped and stored by column in Compression Units (CU), automatically applied during Direct Path Load/Insert operations.

TCC supports four configurable compression levels—Query Low, Query High, Archive Low, and Archive High—allowing administrators to fine-tune performance and storage efficiency according to workload characteristics.

With TCC, users can achieve up to 10x data size reduction, significantly lowering storage costs while maintaining read performance.

Option	Compression Level	Use Case
Query Low	Low	Balanced compression for high-frequency transaction processing
Query High	Medium	Ideal for analytical queries and general data retrieval
Archive Low	High	Prioritizes storage efficiency for archival data
Archive High	Maximum	Designed for read-only, long-term data archives

2. Bitmap Build Optimization in Direct Path (DP) Loading

Performance for massive data loading in cluster environments (TAC) has been optimized through refined bitmap build mechanisms.

Lock Elimination

When Tibero determines that target blocks are not in a dirty or current state across DB instances (e.g., blocks of previously deleted segments), unnecessary lock acquisitions are skipped, minimizing synchronization overhead.

Memory Efficiency & Throughput Gains

Selective locking reduces memory usage and improves throughput, enabling faster and more scalable parallel data ingestion.

3. Multi Redo Buffer — Minimizing Redo Log Contention

Previously, each database instance was limited to a single Redo Log Buffer, often causing contention under high DML concurrency.
With Multi Redo Buffer, Tibero now supports multiple redo buffers per instance, effectively distributing load across sessions and significantly enhancing DML throughput in multi-session environments.

4. Accelerated PL/SQL Execution

A new shared-memory caching mechanism for PL/SQL binary code has been introduced.

This eliminates the need for repetitive Data Dictionary lookups during each call, dramatically improving performance for triggers and user-defined functions (UDFs) in high-frequency execution environments.

5. Pending Statistics — Enhanced Statistics Management

Tibero now supports Pending Statistics, allowing administrators to collect and store database statistics in a deferred (non-applied) state.
These statistics can be tested before being published, ensuring stable query performance.
Pending statistics can also be exported and imported across databases, simplifying SQL tuning and cross-environment optimization.

At TmaxTibero, every patchset release reflects direct feedback from customers and ongoing efforts to enhance performance, stability, and usability.
Detailed documentation on additional feature improvements and configuration guidelines is available through the Tibero TechNet manuals.