Scalability and Performance
Enhance horsepower
|
-
General plan cache
Improve general read performance
-
Partitioned Raft KV storage engine
Provide increased write velocity, faster scaling operations,
larger clusters
-
TiFlash performance boost
Implement TiFlash optimization such as late materialization
and runtime filter
-
Fastest online DDL distributed framework
Complete the distributed framework to support the fastest
online DDL
|
-
Stability at PB scale
Provide reliable and consistent performance for tremendous
data
-
Disaggregate TiFlash compute and storage (auto-scaling)
Realize elastic HTAP resource utilization
-
TiFlash S3 based storage engine
Provide shared storage at lower cost
|
-
Next generation, more powerful storage engine
-
Unlimited transaction size
-
Multi-model support
|
Reliability and Availability
Enhance dependability
|
-
Resource control: quotas and scheduling for resource groups
and background tasks
Reliably and efficiently manage workloads and applications
that share the same cluster
-
TiCDC/PITR recovery objective enhancements
Increase business continuity and minimize the impact of system
failures
-
TiProxy
Keep database connections during cluster upgrade and scale
in/out, and avoid impact on applications
-
End-to-end data correctness check
Prevent data error or corruption through TiCDC
|
-
Multi-tenancy
Provide fine-grained resource control and isolation to reduce
cost
-
Improved cluster/node level fault tolerance
Enhance cluster resilience
-
TiFlash spill to disk
Avoid TiFlash OOM
|
-
Enhanced TiDB memory management
-
Global table
|
SQL
Enhance functionality and compatibility
|
-
Production-ready TTL (time-to-live) data management
Manage database size and improve performance by automatically
expiring outdated data
-
Table level flashback
Support traveling a single table to a specific time via SQL
|
-
Materialized views
Support pre-calculation to boost query performance
|
-
Federated query
-
Cascades optimizer
-
Full text search & GIS support
-
User-defined functions
|
Database Operations and Observability
Enhance DB manageability and its ecosystem
|
-
Distributed TiCDC single table replication
Dramatically improve TiCDC throughput by distributing the
workload to multiple nodes
-
Production-ready TiCDC sink to Amazon S3 and Azure object
storage
Enhance ecosystem to better work with big data
-
TiDB Operator fast scale-in
Improve from scaling in one by one to scaling in at once
-
SQL-based data import
Improve user-friendliness through operational enhancements
|
-
Major performance boost for data import
Expect 3-4 times of improvements
-
Multiple upstreams for TiCDC
Support N:1 TiDB to TiCDC
-
SQL-based data management
Improve data management for TiCDC, data migration, and backup
and restore tools
-
Automatic pause/resume DDL during upgrade
Ensure a smooth upgrade experience
|
-
AI-indexing
-
Heterogeneous database migration support
-
Re-invented AI-SQL performance advisor
|
Security
Enhance data safety and privacy
|
-
JWT authentication
Provide secure and standard authentication
-
LDAP integration
Authenticate via LDAP server over TLS
-
Audit log enhancement
Enhance logs with greater details
|
-
Column-level/row-level access control
Provide finer-grained control
-
Database encryption
Support data-at-rest encryption for database files
-
Unified TLS CA/Key rotation policy
Enhance security and operational efficiency for all TiDB
components
|
-
Enhanced client-side encryption
-
Enhanced data masking
-
Enhanced data lifecycle management
|
