4.1.1
Enhancement
- Add Sharding-Scaling & Sharding-UI dockerfile
- update MySQL & PostgreSQL version for proxy
Bug Fixes
- Fix parser syntax rule of SUBSTRING and CONVERT
- Fix parser ColumnSegment ClassCastException
- Fix TableMetaData load error when use Sharding-JDBC with oracle
- Fix getSchema NPE when use Sharding-JDBC with oracle
- Fix Sharding-JDBC parse sql NPE in PostgreSQL
- Fix Sharding-Proxy receive error response for PostgreSQL JDBC client
- Fix Sharding-Proxy response number of update is 0 for PostgreSQL JDBC client
- Fix Sharding-Proxy receive null for PostgreSQL column meta data
- Fix Sharding-Scaling NPE in MySQL incremental task
Change Logs
- MILESTONE
4.1.0
New Features
- Support scaling for ShardingSphere (alpha version)
- Move etcd registry center from experiment repo to apache repo
- Upgrade to Java 8
Enhancement
- Optimization for Sharing Parser with ANTLR Visitor improving the parsing performance of long SQL by 100%~1000%
- Use multiple threads to load metadata for different data sources
- Support
allow.range.query.with.inline.sharding
option - The docker of ShardingSphere-Proxy supports loading external lib packages
- Support integration with Spring using @ShardingTransactionType
- Enhance ShardingDataSource to compatible with Seata in micro-service distribution transaction
Refactor
- Remove leaf key generator
Bug Fixes
- Fix an exception caused by using a CHAR/VARCHAR type column as an order by item
- Refine
DataTypeName
syntax rules of all database dialects - Fix an exception caused by executing
BEGIN
using prepared statement of MySQL C API - Fix the problem that
ALTER TABLE
fails to execute when the field type of the table contains Integer/Double/BigDecimal - Fix the problem of the stop index dislocation of segment with alias
- Fix the problem that overwriting SQL
SELECT * FROM tbl WHERE encrypt_col = ? AND (sharding_key = ? OR sharding_key = ?)
when using sharding + encrypt would throw StringIndexOutOfBoundsException - Fix the problem of incorrect decoding after AES encoding when using ShardingSphere-Proxy in Spring Boot
- Fix a long-time problem of adding schema dynamically in ShardingSphere-UI
Change Logs
- MILESTONE
4.0.1
Bug Fixes
- Using guava cache to fix parsing deadlock.
- Oracle insert SQL could not work in encrypt mode.
- Proxy for PostgreSQL decode parameters error in all types except String.
- COM_STM_EXECUTE of proxy for MySQL could not support sysbench.
- None sharding strategy could not config in spring-boot.
- Plain column could not get from resultSet in encrypt mode.
- WasNull field was wrong in GroupByStreamMergeResult.
- Metadata.getColumns could not work in JDBC.
- IN operator contains space and
\n
\t
\r
could not supported by parser.
Enhancement
- Optimize antlr performance using two-stage parsing strategy.
- Add class filter constructor to restrict the illegal class from YAML.
Change Logs
- MILESTONE
4.0.0
API Changes
- Change package and maven groupId form
io.shardingsphere
to org.apache.shardingsphere
. - Adjust ShardingSphere-JDBC configuration API.
- Adjust persist structure for registry center.
New Features
- SQL92 Syntax available.
- ShardingSphere-Proxy for PostgreSQL protocol available.
- SQL 100% compatible if route to single data node.
- Less-than(<), greater-than(>) and Less-than-equal(<=), greater-than-equal(>=) for sharding key operator available.
- DISTINCT SQL syntax available.
- Broadcast table available.
- LEAF key generator available.
- XA Transaction available, Atomikos, Narayana and Bitronix integrated.
- BASE Transaction available, Seata integrated.
- Data encrypt available.
- Skywalking plugin available.
- ShardingSphere-UI available, an orchestration management platform.
Enhancement
- MariaDB supported.
- Improve the compatibility of SQL parsing.
SELECT FOR UPDATE
route to primary data source only.- Hint in ShardingSphere-Proxy available.
- Make configuration of orchestration consistent between ShardingSphere-JDBC and ShardingSphere-Proxy.
- Renew modified data sources only, not renew all the data sources.
- Vibrate configurable for Snowflake key generator.
Bug Fixes
- Improve the compatibility of JDBC Driver URL.
- Delete statement with alias available.
- Check and disable updating sharding column.
- Fix wrong type of TINYINT and SMALLINT as INTEGER.
Change Logs
- MILESTONE #3
- MILESTONE #4
- MILESTONE #5
- MILESTONE #6
- MILESTONE #7
- MILESTONE #8
- MILESTONE #9
4.0.0.RC3
New Features
- ShardingSphere-UI, an orchestration management platform for ShardingSphere comes online.
- Not only SQLs from MySQL, PostgreSQL, SQLServer, Oracle, but any SQL92 Syntax can be parsed correctly and used in ShardingSphere.
Enhancement
- Support using less-than character(<) and greater-than character(>) for sharding data.
- When primary and replica dataSources exist, support executing
SELECT FOR UPDATE
on primary data source. - Support hint in ShardingSphere-Proxy.
- Finish parsing DAL syntax for MySQL.
- Make configuration of orchestration compatible between ShardingSphere-JDBC and ShardingSphere-Proxy.
Bug Fixes
- Through Bug fix, the feature of encryption becomes much stable and applicable.
- Support delete statement with alias.
- Check and disable updating sharding column.
- Fix wrong type of TINYINT and SMALLINT as INTEGER.
Refactor
- Rename optimized module to preprocessor module.
- Decouple rewrite core module and sharding/encrypt features.
Change Logs
- MILESTONE
4.0.0.RC2
API Changes
- Optimize and delete API and configuration item of sharding logic index.
- Update the API of encryption to support the encrypted and plain data coexistence.
New Features
- Integration of Seata for distributed transaction.
- User can do data encryption by using ShardingProxy.
- User can use Leaf-segment generator to get distributed ID.
- Support Skywalking plugin for application performance management.
Enhancement
- Renew modified dataSources, not all the datasources to improve performance for configuration orchestration.
- Improve the compatibility of SQL parsing.
Refactor
- Remove DatabaseType enum, use dynamic SPI to load data source type.
- The parse engine upgrade from the 2nd generation to 3rd.
- The Refactoring of SQL rewriting module.
Change Logs
- MILESTONE
4.0.0.RC1
Merge all change logs of version 3.1.0.M1, 3.1.0, 3.1.0.1 and 4.0.0.M1. First apache release.
API Changes
- Adjust persist structure for orchestration's registry center.
- Adjust ShardingSphere-JDBC configuration API.
- Change package and maven groupId form
io.shardingsphere
to org.apache.shardingsphere
. - Adjust spring-boot-starter.
New Features
- XA Transaction available.
- Data encrypt available.
- Use PostgreSQL protocol access ShardingSphere-Proxy available.
- DISTINCT SQL syntax available.
- Broadcast table.
- All SQL 100% compatible if route to single data node (MySQL Only).
Change Logs
- MILESTONE #3
- MILESTONE #4
- MILESTONE #5
- MILESTONE #6
3.0.0
Milestones
- ShardingSphere-Proxy launch. Support the use of ShardingSphere in the form of database to support for MySQL CLI and GUI client
New Features
Core
- ISSUE #290 Support batch INSERT
- ISSUE #501 Support OR
- ISSUE #980 Support DCL
- ISSUE #1111 Support MySQL DAL
ShardingSphere-Proxy
- ISSUE #902 Support XA transaction
- ISSUE #916 Support authorization
- ISSUE #936 Support registry center
- ISSUE #1046 Support multiple logic databases
Enhancements
Core
- ISSUE #373 Support
order by ?
- ISSUE #610 Route unicast for DQL without table
- ISSUE #701 Caching parsed results to improve performance
- ISSUE #773 Support sharding and autoincrement key of INSERT without column names
- ISSUE #935 Use
YAML
instead of JSON
to store configurations in registry center - ISSUE #1004 Properties can configure for sharding and primary-replica replication independent
- ISSUE #1205 Execute engine enhancement
ShardingSphere-JDBC
- ISSUE #652 Support
Spring Boot Starter
2.X - ISSUE #702 Support
$->{..}
for inline expression - ISSUE #719 Support inject key generator objects by spring namespace
- ISSUE #720 Support inject sharding algorithm objects by spring namespace
Sharding-Opentracing
- ISSUE #1172 Opentracing enhancement
API changes
- ISSUE #1153 Adjust the maven artifactId for Orchestration module
- ISSUE #1203 Adjust Spring namespace xsd for Sharding and primary-replica replication
- ISSUE #1289 Adjust Hint API
- ISSUE #1302 Refine package structure
- ISSUE #1305 Deprecated and remove sharding-jdbc-transaction-parent module
- ISSUE #1382 Remove type configuration in Orchestration module
Bug Fixes
Core
- ISSUE #569 Failed to parse SQL for Oracle when ROWNUM is not at end
- ISSUE #628 Support data type jsonb for PostgreSQL
- ISSUE #646 When aliases in
SELECT ITEMS
correspond to the real column names of GROUP BY
or ORDER BY
, there is no need to generate derived columns - ISSUE #806
NOT IN
parse exception - ISSUE #827 Endless loop for bad SQL like
SELECT * FROM table WHERE id IN ()
- ISSUE #919 Inline expression with groovy may cause memory leak
- ISSUE #993 Fail to parsing PostgreSQL due to the quotation
- ISSUE #1015 Support SQL like
SELECT id, COUNT(*) FROM table GROUP BY 1,2
- ISSUE #1120 Derived columns of
GROUP BY / ORDER BY
appear in query result - ISSUE #1186 Dead lock may occur on MEMORY_STRICTLY mode when get connection on concurrency environment
- ISSUE #1265 RoundRobinReplicaLoadBalanceAlgorithm throw an ArrayIndexOutOfBoundsException when AtomicInteger overflow
ShardingSphere-JDBC
- ISSUE #372 Reuse PreparedStatement cause cache of route result do not clean
- ISSUE #629 Support transaction isolation on JDBC
- ISSUE #735 Unexpected replica datasource routing result when using
Round-robin
load-balance algorithm in Mybatis - ISSUE #1011 Can't resolve placeholder in
Spring Boot YAML
configuration
2.0.3
New Features
Core
- ISSUE #600 Support TCL
Bug Fixes
Core
- ISSUE #540 Support SQL that alias is the keyword
- ISSUE #577 Support new line for
YAML
configuration
ShardingSphere-JDBC
- ISSUE #522 Replica database does not need to execute the DDL for primary-replica replication
2.0.2
Enhancements
Core
- ISSUE #475 Support
CREATE INDEX
- ISSUE #525 Support
DROP INDEX
Bug Fixes
Core
- ISSUE #521
ShardingProperties
is invalid in YAML
configuration - ISSUE #529 Table name capitalization cannot be queried
- ISSUE #541
IS NOT NULL
parse error - ISSUE #557 When
GROUP BY
and ORDER BY
aliases are inconsistent, stream merging should be used - ISSUE #559 Support parsing numbers beginning with minus and decimal (e.g.
-.12
) - ISSUE #567 Add escape char for derived columns or alias when using MySQL
ShardingSphere-JDBC
- ISSUE #520 Exception is no longer
DuplicateKeyException
when the unique key conflict
2.0.1
Enhancements
Core
- ISSUE #490 Using
rownum
greater than or equal to or less than or equal to the result of paging is incorrect in Oracle - ISSUE #496 Logical table names in sharding configuration can be case sensitive
- ISSUE #497 Close registry center gracefully
Bug Fixes
ShardingSphere-JDBC
- ISSUE #489 Uses
RuntimeBeanReference
to prevent the creation of InnerBean
on spring namespace - ISSUE #491 Can't close connection by
ResultSet.getStatement().getConnection().close()
2.0.0
Milestones
- API adjust. Brand new groupId and artifactId for
Maven
, package name and spring namespace name. Simplify and enhance API configuration, inline expression fully configuration support - Support
spring-boot-starter
of ShardingSphere-JDBC
- Dynamic configuration.
ZooKeeper
and etcd
can be used as registry to dynamically modify data sources and sharding configurations - Database orchestration. Fusing database access procedures to access databases and disable access to replica databases
- ConfigMap support. Predefined metadata can be obtained in the sharding and primary-replica replication strategy
- Tracking system support. You can view the invocation chain of
ShardingSphere-JDBC
through sky-walking
and other Opentracing
based APM systems
Enhancements
Core
- ISSUE #386 Support SQL that does not contain table names, such as
SELECT 1
ShardingSphere-JDBC
- ISSUE #407 Support Hyphen properties for
sharding-jdbc-spring-boot-starter
- ISSUE #424 Providing SQL overall execution events
Bug Fixes
Core
- ISSUE #387 Prevent errors from keywords process when ‘`’ exists in function + column name
- ISSUE #419 When SQL is rewritten, it does not determine whether alias is a keyword without the escape character, which results in SQL exception
- ISSUE #464 SQL if the varchar type is not closed due to the absence of matching single quotes, and the next varchar in SQL is the wrong SQL of Chinese characters, it will lead to higher use of CPU
ShardingSphere-JDBC
- ISSUE #394 Can't only close statement
- ISSUE #398 Use Hint routing to shield case sensitivity
- ISSUE #404 Sharding-jdbc's spring-boot-starter does not support HikariDataSource
- ISSUE #436 Primary-replica replication, when the RoundRobin algorithm is configured from the database and MyBatis is used, it can only be routed to the same replica database
- ISSUE #452 Sharding of DDL statements to more than one table causes a connection leak
- ISSUE #472 Before Connection executes createStatement, it calls getMetaData first and then setAutoCommit can not take effective connection to the database that was created later
1.5.4.1
Bug Fixes
- ISSUE #382 The query cannot be completed without sharding rule
1.5.4
Bug Fixes
- ISSUE #356 In the Where condition of SQL, the REGEXP operator is compatible with non sharding columns
- ISSUE #362 Primary-replica replication using PreparedStatement does not invoke the setParameter method to cause errors
- ISSUE #370 Error in calling getGeneratedKeys using native self increment primary key
- ISSUE #375 Data can not be obtained after paging second pages route to a single node
- ISSUE #379 When Mybatis is used to call Connection.getMetaData (), the connection is not close correct
1.5.3
Enhancements
- ISSUE #98 Primary-replica replication load balancing strategy support configuration
- ISSUE #196 Primary-replica replication and sharding configuration independence
Bug Fixes
- ISSUE #349 Incorrect function of ResultSet.wasNull causes null numeric type in DB to zero
- ISSUE #351 Tables that are included in the default data source but not in TableRule configuration are not properly executed
- ISSUE #353 In the Where condition of SQL, it is compatible with non sharding columns !=, !> and !< operator
- ISSUE #354 In the Where condition of SQL, NOT operators are compatible with non-sharding columns
1.5.2
Milestones
- The test engine of quality assurance, each SQL can run 60 test cases of different dimensions
Enhancements
- ISSUE #335 Support the GROUP BY + custom function SQL
- ISSUE #341 Support ORDER BY xxx NULLS FIRST | LAST statement of Oracle
Bug Fixes
- ISSUE #334 Parsing ORDER BY with functions will resolve the following ASC and DESC to the name attribute of OrderItem
- ISSUE #335 JOIN parsing is incorrect using the full name association of the table
- ISSUE #346 Parsing table name error of DDL statement DROP TABLE IF EXISTS USER
1.5.1
New Features
- ISSUE #314 Support DDL type SQL
Changes
- ISSUE #327 Close sql.show configuration by default
Bug Fixes
- ISSUE #308 Invalid return of database native GeneratedKey
- ISSUE #309 ORDER BY and GROUP BY in sub-queries are not included in the analytic context
- ISSUE #313 Support <> operator
- ISSUE #317 Parameter of LIMIT can not be type of Long
- ISSUE #320 SQL rewriting error of GROUP BY + LIMIT
- ISSUE #323 Parsing ORDER BY + Aggregation Expression error
1.5.0
Milestones
- The new SQL parsing module removes the dependence on Druid. We only need to parse the sharding context, and adopt a “semi understanding” concept for SQL to further improve performance and compatibility, and reduce code complexity
- The new SQL rewrite module adds an optimized rewrite module
- The new SQL merge module is rebuilt into 3 merging engines: streaming, memory and decorator
New Features
- Support Oracle, SQLServer and PostgreSQL
- Non functional sub-query support
Enhancements
- ISSUE #256 Configurable display sharding execute to SQL log
- ISSUE #291 Processing SQL use stream mode that contains only GroupBy
Changes
- Simplify the distributed self increasing sequence. Each table is supported by simplifying a multiple self increasing sequence to a single table supporting only a single distributed self increasing sequence, and no longer supporting workerID settings through environment variables
- Remove support for OR
Bug Fixes
- ISSUE #239 LIMIT routes to multiple query result sets. If there is only one result set that is not empty, the result of paging is incorrect
- ISSUE #263 Sharding and logical table configuration can be case insensitive
- ISSUE #292 When the memory mode handles GROUP BY statement, if there is paging information, it needs to be rewritten
- ISSUE #295 LIMIT 0 does not filter the result set according to paging restrictions
1.4.2
Enhancements
- ISSUE #219 Thread performance optimization
- ISSUE #215 Aggregated result set of stream sort StreamingOrderByReducerResultSet performance optimization
- ISSUE #161 When the result sets are merged, heap sort can be used to improve performance
Bug Fixes
- ISSUE #212 More meaningful hints for missing data source rules
- ISSUE #214 table_name.column_name in (?,?) in WHERE can't parser expression
- ISSUE #180 Batch execution of Update return inaccuracy
- ISSUE #225 The last character of automatic generation of Id is not 0
1.4.1
Enhancements
- ISSUE #191 Generating KeyGenerator of workerId based on IP of host
- ISSUE #192 Get workerId‘s KeyGenerator based on HOSTNAME’s digital suffix
- ISSUE #210 Routing to single database and single table to remove supplementary SQL statement fragments
Bug Fixes
- ISSUE #194 Some component exceptions in the close method of Connection, Statement, ResultSet and other interfaces cause the close method of another component to be not invoked
- ISSUE #199 Sharding and reuse PreparedStatement object cause route error
- ISSUE #201 Event transmission missing before batch operation execution
- ISSUE #203 Merge events sent by the batch operation
- ISSUE #209 Parallel execution of multiple limit queries leads to IndexOutOfBoundsException
1.4.0
Enhancements
Automatic generation key implementation, including
- ISSUE #162 Implementation of distributed primary key algorithm
- ISSUE #163 Acquisition of a self increasing sequence of JDBC interfaces
- ISSUE #171 Sharding-jdbc-core coordination automatic generation sequence transformation
- ISSUE #172 The configuration of YAML and Spring supports the self increasing sequence
Bug Fixes
- ISSUE #176 The wasNull flag of AbstractMemoryResultSet is not reset in time
1.3.3
Enhancements
- ISSUE #59 PreparedStatement can call the correct underlying set method according to the parameter type when setting parameters
Bug Fixes
- ISSUE #149 When INSERT IGNORE INTO, if the data is duplicated, the value returned to -1 when ignored, and it should be returned to 0
- ISSUE #118 In the same thread, DQL is executed first, then DML is executed, and DML operation is executed from the replica database
- ISSUE #122 In cases where connections are not available (such as network interruption), transactions should be interrupted rather than retry
- ISSUE #152 PreparedStatement's cache causes an array out of bound
- ISSUE #150 With the latest SQLServer jdbc driver compatibility problem, Product Name should be changed from SQLServer to Microsoft SQL Server
- ISSUE #166 Druid data source stat filter multi-thread error reporting should be added to database connection level synchronization
1.3.2
Enhancements
- ISSUE #79 Optimizes limit for only one target table, does not modify limit offset
Bug Fixes
- ISSUE #36 ShardingPreparedStatement cannot set parameters repeatedly
- ISSUE #114 When ShardingPreparedStatement performs batch tasks, it repeatedly analyzes SQL and leads to OOM
- ISSUE #33 According to the MySQL document, queries similar to limit 100 and -1 format are not supported
- ISSUE #124 The return value of com.dangdang.ddframe.rdb.sharding.jdbc.adapter.AbstractStatementAdapter.getUpdateCount does not conform to the JDBC specification
- ISSUE #141 Multi-thread executor parameter setting failure
1.3.1
Enhancements
- ISSUE #91 Open support for Statement.getGeneratedKeys can return the original database self increase primary key
- ISSUE #92 Query DQL statement event sending
Bug Fixes
- ISSUE #89 Use primary-replica replication with sharding hint leads to conflict
- ISSUE #95 Write operations in the same thread read from the primary database changed to the same thread and within the same connection
1.3.0
New Features
- ISSUE #85 primary-replica replication separation
Enhancements
- ISSUE #82 TableRule can import the dataSourceName attribute to specify the data source corresponding to the TableRule
- ISSUE #88 Release restrictions on other databases, support standard SQL, do not support personalized paging statements
Bug Fixes
- ISSUE #81 Associative table query uses OR query condition to resolve the result exceptions
1.2.1
Refactor
- ISSUE #60 API adjust, remove ShardingDataSource, use factory instead
- ISSUE #76 ShardingRule and TableRule change to Builder pattern
- ISSUE #77 ShardingRule and TableRule change to Builder pattern
Enhancements
- ISSUE #61 Add the logical table name to the ShardingValue class
- ISSUE #66 Statement on the JDBC tier supports get/set MaxFieldSize, MaxRows and QueryTimeout
- ISSUE #72 Batch inserts supporting select union all
- ISSUE #78 Simplifying sharding only configuration, without configuring logical table and real table correspondence
- ISSUE #80 Simplifying the configuration that does not sharding, specifying the default data source, do not need configure TableRule
Bug Fixes
- ISSUE #63 No table name or table alias is added to the ORDER BY and GROUP BY derivation columns
- ISSUE #65 Performance enhancement for parsing condition context
- ISSUE #67 The soft transaction log cannot be deleted when routed to multiple tables
- ISSUE #71 Routing single sharding key by OFFSET of LIMIT error
- ISSUE #75 MemoryTransactionLogStorage retry times update concurrency problem
1.2.0
New Features
- ISSUE #53 The relationship between the real table and the logical table is not configured, and the real table is dynamically calculated by the sharding algorithm
- ISSUE #58 Soft transaction: the initial version of the best effort type
Refactor
- ISSUE #49 Adjust the property configuration
- ISSUE #51 Refactor of Hint interface
Bug Fixes
- ISSUE #43 The yaml file contains Chinese, and the operating system mode is not UTF-8 encoding, resulting in yaml can not be parsed
- ISSUE #48 Yaml file is not closed after reading
- ISSUE #57 At the analytic level, we can identify subqueries to ensure that the behavior of supplementary columns can be accurately positioned
1.1.0
New Features
- ISSUE #40 Support YAML configuration
- ISSUE #41 Support Spring namespace configuration
- ISSUE #42 Support inline expression configuration
Bug Fixes
- ISSUE #25 The problem of repeated results will appear under the OR expression
1.0.1
Enhancements
- ISSUE #39 Support the use of Hint method to register the key value to SQL routing
Bug Fixes
- ISSUE #11 The count function returns incorrectly without aliases
- ISSUE #13 The Insert statement does not write column names or write column names but column names do not contain sharding fields, occur broadcast route
- ISSUE #16 For now a new connection pool is executed every time SQL is executed. Instead, each ShardingDataSource object should be changed to share a pool
- ISSUE #18 When query Count, getObject() throws Exception: Unsupported data type: Object
- ISSUE #19 In SUM and AVG functions, merger is not executed if aliases are not added, and null pointer exception fired if aliases are added
- ISSUE #38 The compatibility between JPA and ShardingSphere-JDBC. JPA automatically add the column aliases of SELECT, resulting in ORDER BY obtaining ResultSet data only by aliases rather than column names
1.0.0
- Initial version