Key Features of 2016 RDBMS Releases: Oracle, MySQL, PostgreSQL, SQL Server, DB2 and More

Oracle 12c Release 2

Oracle Database 12c Release 2 documentation was made public in mid‑November 2016. Compared with Release 1, the new release strengthens Multitenant, In‑Memory, Spatial & Graph, and Sharding capabilities.

Core Database : a new Linux‑only package installable via yum install oracle-core-database, licensed but fully supported.

Sharding : partitions can reside in separate databases, managed by a Shard Directory and a Coordinator DB when a Sharding Key is not supplied.

Global Data Services (GDS) : provides a connection‑pool proxy for routing Shard Directory entries.

Multitenant enhancements : supports up to 4,096 PDBs, Hot Clone, Refresh, and online tenant migration.

In‑Memory Option (IMO) : improved read/write separation, automatic hot‑data migration, and dynamic cold‑data eviction.

DataGuard improvements : simplified backup creation via DBCA, automatic password‑file replication, remote AWR snapshots, connection‑session keepalive, and expanded automatic block repair.

Parallel redo apply : multi‑instance parallel recovery for DataGuard.

BIG SCN : SCN size increased from 6 to 8 bytes, raising the maximum SCN value.

Cross‑Endian Dictionary : enables Oracle data dictionaries to be read/written across big‑ and little‑endian platforms.

Other tweaks : V$SQL now shows redo size per statement; object name length raised to 128 characters; automatic index updates during table moves.

MySQL 5.7

MySQL 5.7 introduces a range of security and functional enhancements.

Mandatory non‑null plugin column in mysql.user with default mysql_native_password; mysql_old_password removed.

Password expiration controlled by default_password_lifetime.

User account lock/unlock via ALTER USER ... ACCOUNT LOCK.

Default SSL encryption for client connections.

New initialize method replaces mysql_install_db, creating only a temporary root account and no test database.

Native JSON data type with binary storage and a rich set of JSON functions (creation, modification, query, and property functions).

New sys schema provides easy‑to‑read performance and resource views.

In‑place ALTER for VARCHAR size changes (subject to byte‑length constraints).

Online DDL for tables and partitions (e.g., OPTIMIZE TABLE, ALTER TABLE … FORCE).

Support for DATA_GEOMETRY spatial type.

Enhanced innodb_buffer_pool_dump_pct for selective buffer‑pool dumps.

Multi‑threaded dirty‑page flushing controlled by innodb_page_cleaners (max 64).

Dynamic buffer‑pool resizing without restart (chunk‑based migration).

Undo‑log truncation via innodb_undo_log_truncate.

Built‑in full‑text parser supporting Chinese, Korean, Japanese.

Online replication filter adjustments, multi‑source replication, and group‑commit parallel replication.

GTID enhancements: online enable/disable and storage in mysql.gtid_executed.

Percona Server 5.7.16‑10

Percona Server 5.7.16‑10, based on MySQL 5.7.16, adds performance and stability improvements.

Deprecated Metrics for scalability measurement plugin removed.

Bug fixes for slow‑admin statement logging, crash recovery with innodb_force_recovery=6, thread‑pool overflow handling, and TokuDB statistics accuracy.

Additional bug fixes: #1633061, #1633430, #1635184.

MariaDB 10.1.19

MariaDB 10.1.19, a stable branch of MySQL, introduces several updates.

XtraDB upgraded to 5.6.33‑79.0.

TokuDB upgraded to 5.6.33‑79.0.

Ubuntu 16.10 package support.

Security fixes for CVE‑2016‑7440 and CVE‑2016‑5584.

SQL Server 2016 SP1

SQL Server 2016 SP1 adds enterprise‑grade features to all editions.

Row‑level encryption, dynamic data masking, change tracking, database snapshots, columnstore indexes, table partitioning, table compression, In‑Memory OLTP, Always Encrypted, PolyBase, audit, multiple filestream containers.

Database cloning via DBCC CLONEDATABASE.

CREATE OR ALTER syntax for stored procedures, triggers, functions, and views.

USE HINT query hint syntax.

Showplan XML enhancements (memory grant warnings, CPU time, wait stats).

New DMF sys.dm_db_incremental_stats_properties.

EstimatedRowsRead attribute in execution plans.

Automatic TABLOCK hint for bulk inserts on heap tables (trace flag 715).

Reduced Hekaton error‑log noise.

SQL Server on Linux (vNext CTP1)

Technical preview of SQL Server for Linux (vNext CTP1, internal version 14.0) demonstrates component‑based installation and a 138 MB core engine package, with full documentation available on Microsoft Docs.

PostgreSQL 9.6

PostgreSQL 9.6, released 29 September 2016, brings major improvements.

Parallel query execution across multiple cores.

Strong‑consistent multi‑replica synchronous replication with configurable quorum.

Remote‑apply synchronous level for strict read‑write consistency.

Full‑text search enhancements (phrase matching, spelling dictionary).

postgres_fdw now pushes JOIN, SORT, UPDATE, DELETE to remote servers.

High‑concurrency performance (up to 1.8 M tps on a 72‑core node).

VM freeze‑bit recognition to speed up static data scans.

69 new wait‑event counters for bottleneck analysis.

Improved sorting kernel leveraging CPU cache.

Snapshot timeout handling.

IO scheduler optimizations (checkpoint ordering, smoother background writer activity).

Aggregation function reuse for identical operators.

Bloom filter support for arbitrary column indexes.

PostgreSQL 9.6.1 Patch

Released 27 October 2016, the 9.6.1 patch primarily provides bug fixes.

DB2 LUW V11.1

IBM released DB2 LUW 11.1 on 15 June 2016, adding extensive enhancements.

Column‑organized table support for partitioned environments (MPP‑aware query plans, vector‑format data exchange, public table dictionary for compressed data across networks).

Column‑store tables with NOT LOGGED INITIALLY for faster DML.

Automatic Dictionary Creation (ADC) during bulk import for higher compression.

Improved high‑concurrency performance (reduced latch contention, faster transaction latency).

Row‑ and column‑level access control (RCAC) for column‑store tables.

Enterprise key management integration (KMIP‑compatible, HSM support).

PureScale enhancements: synchronous HADR, unified workload management, GDPC dual‑active improvements, GPFS replication via db2cluster, and health‑check commands.

Greenplum 4.3.10.0, GPCC 3.0, GPText 2.0

Greenplum, an MPP‑based PostgreSQL fork, added S3 writable tables and Azure cloud support in October 2016.

GPCC 3.0 : HTML5 UI, responsive design, Go‑based RESTful API, reduced resource usage, test‑driven development, CI/CD pipeline.

GPText 2.0 : integrates SolrCloud 6.1, high‑availability features, and Madlib analytics; exposes UDFs for text indexing and search via SQL.

Additional performance and feature upgrades (ORCA optimizer, gpfdist, gpcheckcat, gpcrondump).

All listed releases were compiled from official documentation and vendor announcements.