PolarDB-X is a cloud native distributed SQL Database designed for high concurrency, massive storage, complex querying scenarios.

Last update: Dec 31, 2022

Overview

What is PolarDB-X ？

PolarDB-X is a cloud native distributed SQL Database designed for high concurrency, massive storage and complex querying scenarios. It has a shared-nothing architecture in which computing is decoupled from storage. It supports horizontal scaling, distributed transactions and Hybrid Transactional and Analytical Processing (HTAP) workloads, and is characterized by enterprise-class, cloud native, high availability, highly compatible with MySQL and its ecosystem.

PolarDB-X was originally created to solve the database's scalability bottleneck of Alibaba Tmall's "Double Eleven" core transaction system, and has grown with AliCloud along the way, and is a mature and stable database system that has been verified by many customers' core business systems.

The core features of PolarDB-X include:

Horizontal Scalability

PolarDB-X is designed with Shared-nothing architecture, supporting multiple Hash and Range data sharding algorithms and achieving transparent horizontal scaling through implicit primary key sharding and dynamic scheduling of data shard.

Distributed Transactions

PolarDB-X adopts MVCC + TSO approach and 2PC protocol to implement distributed transactions. Transactions meet ACID characteristics, support RC/RR isolation levels, and achieve high performance through optimizations such as one-stage commit, read-only transaction, and asynchronous commit.

HTAP

PolarDB-X supports analytical queries through native MPP capability, and achieves strong isolation of OLTP and OLAP traffic through CPU quota constraint, memory pooling, storage resource separation, etc.

Enterprise-class

PolarDB-X has many capabilities designed for enterprise scenarios, such as SQL Concurrency Control, SQL Advisor, TDE, Triple Authority Seperation, Flashback Query, etc.

Cloud Native

PolarDB-X has years of cloud native practice on AliCloud, supports managing cluster resources via K8S Operator, supports public cloud, hybrid cloud, private cloud and other forms for deployment.

High Availability

PolarDB-X achieves strong data consistency through Multi-Paxos protocol, supports cross-data center deployment, and improves system availability through Table Group, Geo-locality, etc.

Compatible with MySQL and Its Ecosystem

The goal of PolarDB-X is to be fully compatible with MySQL, which currently includes MySQL protocol, most of MySQL SQL syntax, Collations, transaction isolation level, binary log, etc.

Quick Start

To quick start with PolarDB-X

PolarDB-X supports one-click installation by PXD tool, through which you can quickly try the functions of PolarDB-X.

See the PXD Quick Start.

To quick start with PolarDB-X on Kubernetes

PolarDB-X provides K8S deployment mode, through which you can customize the configuration of PolarDB-X cluster.

See the K8S Quick Start.

To start developing PolarDB-X

If you want to compile and install PolarDB-X from source code, or start development, you can refer the Development Guide.

The core features of PolarDB-X community version will be consistent with the commercial version, and more manuals can be found in the documentations of the commercial version. The documentations of the community version are being compiled and will be released to the public in the near future.

Architecture

PolarDB-X has a shared-nothing architecture in which compute and storage is decoupled, and the system consists of 4 core components.

CN (Compute Node)

The Compute Node is the entry point of the system and is stateless, which includes modules such as SQL parser, optimizer, and executor. It is responsible for distributed data routing, 2PC coordination, global secondary index maintenance, etc. It also provides enterprise features such as SQL concurrency control and triple authority separation.

DN (Data Node)

The Data Node is responsible for data persistence, providing strong data consistency based on the Multi-Paxos protocol, while maintaining distributed transaction visibility through MVCC.

GMS (Global Meta Service)

The Gloal Meta Service is responsible for maintaining globally consistent Table/Schema, Statistics and other system Meta information, maintaining security information such as accounts and permissions, and providing global timing services (i.e. TSO).

CDC (Change Data Capture)

The CDC Node provides change data capture capability that is fully compatible with the MySQL binary log format and MySQL DUMP protocol, and master-slave replication capability that is compatible with the MySQL Replication protocol.

PolarDB-X provides tool to manage the above components through K8S Operator, and the RPC between the CN and DN can be done through private protocol component. The corresponding repositories of these components are as follows.

Component Name	Repository
CN (Compute Node)	galaxysql
GMS (Global Meta Service)	galaxyengine
DN (Data Node)	galaxyengine
CDC (Change Data Capture)	galaxycdc
RPC	galaxyglue
K8S Operator	galaxykube

What is ApsaraDB GalaxySQL ？

ApsaraDB GalaxySQL is one component of PolarDB-X, namely CN (Compute Node).

Licensing

ApsaraDB GalaxySQL is under the Apache License 2.0. See the License file for details.

Contributing

You are welcome to make contributions to PolarDB-X. We appreciate all the contributions. For more information about how to start development and pull requests, see contributing.

Community

You can join these groups and chats to discuss and ask PolarDB-X related questions:

DingTalk Group: 32432897
WeChat Group: 阿里云 PolarDB-X 开源交流群 (Contact group manager to get into wechat group. Managers' ID: oldbread3, hustfxj, agapple0002)

Acknowledgements

ApsaraDB GalaxySQL references from many open source projects, such as Calcite, Presto etc. Sincere thanks to these projects and contributors.

Comments

【PXC-9001】CN启动失败, 报错: “No found any leader for metadb storage”

操作路径：源码编译DN启动后 1 初始化执行能够正常完成 bin/startup.sh -I -P asdf1234ghjk5678 -r "admin" -d 127.0.0.1:4886:32886 -u polardbx_root -S "123456" 2 把 metaDbPasswd=BVUB3NoOBdCXPv+X2Wa3es6Ng1tzChSd9Vzdm2RVaBo= 放到 conf/server.properties 3 启动命令没有报错 bin/startup.sh -P asdf1234ghjk5678
4 连接数据库报错了，8527端口没有监听 mysql -h127.0.0.1 -P8527 -upolardbx_root -p123456 ERROR 2003 (HY000): Can't connect to MySQL server on '127.0.0.1' (111)

配置如下：

报错日志：

opened by Cleverdada 11
I ran benchmark but got trapped in a trouble, error message would be pasted followed.

alter table bmsql_district add constraint d_warehouse_fkey foreign key (d_w_id) references bmsql_warehouse (w_id); [134bc68428001000][192.168.10.221:10413][db_test]Failed to execute the DDL task. Caused by: ERR-CODE: [PXC-4636][ERR_DDL_JOB_ERROR] Not all physical DDLs have been executed successfully: 1 expected, 0 done, 1 failed. Caused by: ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_warehouse' ;ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_warehouse' . alter table bmsql_customer add constraint c_district_fkey foreign key (c_w_id, c_d_id) references bmsql_district (d_w_id, d_id); [134bc6855f801000][192.168.10.221:10413][db_test]Failed to execute the DDL task. Caused by: ERR-CODE: [PXC-4636][ERR_DDL_JOB_ERROR] Not all physical DDLs have been executed successfully: 1 expected, 0 done, 1 failed. Caused by: ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_district' ;ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_district' . alter table bmsql_history add constraint h_customer_fkey foreign key (h_c_w_id, h_c_d_id, h_c_id) references bmsql_customer (c_w_id, c_d_id, c_id); [134bc68699001000][192.168.10.221:10413][db_test]Failed to execute the DDL task. Caused by: ERR-CODE: [PXC-4636][ERR_DDL_JOB_ERROR] Not all physical DDLs have been executed successfully: 1 expected, 0 done, 1 failed. Caused by: ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_customer' ;ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_customer' . alter table bmsql_history add constraint h_district_fkey foreign key (h_w_id, h_d_id) references bmsql_district (d_w_id, d_id); [134bc687cf401000][192.168.10.221:10413][db_test]ERR-CODE: [PXC-4644][ERR_PENDING_DDL_JOB_EXISTS] Another DDL job '1390423191423291392' with operation 'ALTER_TABLE' is pending on db_test.bmsql_history in db_test. Please use SHOW DDL to check it, and then recover or rollback it using RECOVER DDL or ROLLBACK DDL, or just remove it using REMOVE DDL if you confirm that the pending job can be discarded. alter table bmsql_new_order add constraint no_order_fkey foreign key (no_w_id, no_d_id, no_o_id) references bmsql_oorder (o_w_id, o_d_id, o_id); [134bc688d1001000][192.168.10.221:10413][db_test]Failed to execute the DDL task. Caused by: ERR-CODE: [PXC-4636][ERR_DDL_JOB_ERROR] Not all physical DDLs have been executed successfully: 1 expected, 0 done, 1 failed. Caused by: ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_oorder' ;ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_oorder' . alter table bmsql_oorder add constraint o_customer_fkey foreign key (o_w_id, o_d_id, o_c_id) references bmsql_customer (c_w_id, c_d_id, c_id); [134bc68a08001000][192.168.10.221:10413][db_test]Failed to execute the DDL task. Caused by: ERR-CODE: [PXC-4636][ERR_DDL_JOB_ERROR] Not all physical DDLs have been executed successfully: 1 expected, 0 done, 1 failed. Caused by: ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_customer' ;ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_customer' . alter table bmsql_order_line add constraint ol_order_fkey foreign key (ol_w_id, ol_d_id, ol_o_id) references bmsql_oorder (o_w_id, o_d_id, o_id); [134bc68b3f001000][192.168.10.221:10413][db_test]Failed to execute the DDL task. Caused by: ERR-CODE: [PXC-4636][ERR_DDL_JOB_ERROR] Not all physical DDLs have been executed successfully: 1 expected, 0 done, 1 failed. Caused by: ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_oorder' ;ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_oorder' . alter table bmsql_order_line add constraint ol_stock_fkey foreign key (ol_supply_w_id, ol_i_id) references bmsql_stock (s_w_id, s_i_id); [134bc68c77001000][192.168.10.221:10413][db_test]ERR-CODE: [PXC-4644][ERR_PENDING_DDL_JOB_EXISTS] Another DDL job '1390423211413344256' with operation 'ALTER_TABLE' is pending on db_test.bmsql_order_line in db_test. Please use SHOW DDL to check it, and then recover or rollback it using RECOVER DDL or ROLLBACK DDL, or just remove it using REMOVE DDL if you confirm that the pending job can be discarded. alter table bmsql_stock add constraint s_warehouse_fkey foreign key (s_w_id) references bmsql_warehouse (w_id); [134bc68d78801000][192.168.10.221:10413][db_test]Failed to execute the DDL task. Caused by: ERR-CODE: [PXC-4636][ERR_DDL_JOB_ERROR] Not all physical DDLs have been executed successfully: 1 expected, 0 done, 1 failed. Caused by: ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_warehouse' ;ERR-CODE: [PXC-4614][ERR_EXECUTE_ON_MYSQL] Error occurs when execute on GROUP 'DB_TEST_SINGLE_GROUP' ATOM 'dskey_db_test_single_group#pxc_test-dn-0#192.168.10.221-17298#db_test_single': Failed to open the referenced table 'bmsql_warehouse' . alter table bmsql_stock add constraint s_item_fkey foreign key (s_i_id) references bmsql_item (i_id); [134bc68eaf001000][192.168.10.221:10413][db_test]ERR-CODE: [PXC-4644][ERR_PENDING_DDL_JOB_EXISTS] Another DDL job '1390423220942802944' with operation 'ALTER_TABLE' is pending on db_test.bmsql_stock in db_test. Please use SHOW DDL to check it, and then recover or rollback it using RECOVER DDL or ROLLBACK DDL, or just remove it using REMOVE DDL if you confirm that the pending job can be discarded.
documentation

opened by adzfolc 5
[Enhancement] refactor to simplify the judge bottom-join

https://github.com/ApsaraDB/galaxysql/blob/aa0b41ca6db291c0561089b5c25bd4171a969194/polardbx-calcite/src/main/java/org/apache/calcite/rel/rules/JoinCommuteRule.java#L190-L200

I think we don't need to judge it's bottom-join by using size() > 1.

Instead, we can just use left() instance of Join.

opened by jackwener 4
CBOLogicalSemiJoinLogicalJoinTransposeRule.LASSCOM is redundant

This rule is SemiJoin(LogicalJoin(X, Y), Z).

inside LogicalJoin isn't semi/anti join, which means that left must exist project.

So the patterm must be SemiJoin-project-LogicalJoin, this INSTANCE is redundant

opened by jackwener 3
auto模式数据库BROADCAST表创建失败

mysql> show create database sbtest; +----------+----------------------------------------------+ | DATABASE | CREATE DATABASE | +----------+----------------------------------------------+ | sbtest | CREATE DATABASEsbtest/* MODE = 'auto' */ | +----------+----------------------------------------------+ 1 row in set (0.01 sec) `mysql> CREATE TABLE broadcast_tbl(` ` -> id bigint not null auto_increment,` ` -> bid int,` ` -> name varchar(30),` ` -> primary key(id)` ` -> ) BROADCAST;` `Query OK, 0 rows affected (0.86 sec)` mysql> show rule from broadcast_tbl; +----+---------------+-----------+------------------+---------------------+--------------------+------------------+---------------------+--------------------+ | ID | TABLE_NAME | BROADCAST | DB_PARTITION_KEY | DB_PARTITION_POLICY | DB_PARTITION_COUNT | TB_PARTITION_KEY | TB_PARTITION_POLICY | TB_PARTITION_COUNT | +----+---------------+-----------+------------------+---------------------+--------------------+------------------+---------------------+--------------------+ | 0 | broadcast_tbl | 0 | NULL | NULL | 1 | NULL | NULL | 1 | +----+---------------+-----------+------------------+---------------------+--------------------+------------------+---------------------+--------------------+ 1 row in set (0.00 sec) ``
bug

opened by welyss 3

pxd tryout fails

Following https://www.polardbx.com/document , I run pxd tryout, but I got error below:

Digest: sha256:41d4f23930d5bce8d176944a784b6b40c79d97b7faf5423d2077a1ba526b7b8c
Status: Image is up to date for polardbx/polardbx-init:latest
Pull image: polardbx/xstore-tools:latest at 127.0.0.1


latest:Pulling from polardbx/xstore-tools 
Digest: sha256:5c0b19954c18fd2347eee46dffd82c4babf7e57b2e405d8b4ccb4d5a647c19a7
Status: Image is up to date for polardbx/xstore-tools:latest
Processing  [#########---------------------------]   25%    create gms node

Traceback (most recent call last):
  File "/workdir/polar-x/venv/lib64/python3.6/site-packages/docker/api/client.py", line 268, in _raise_for_status
    response.raise_for_status()
  File "/workdir/polar-x/venv/lib64/python3.6/site-packages/requests/models.py", line 953, in raise_for_status
    raise HTTPError(http_error_msg, response=self)
requests.exceptions.HTTPError: 500 Server Error: Internal Server Error for url: http+docker://localhost/v1.30/containers/9175e58605d9cc143ee42c1e14f7f22b945f09b0fcc7f041e348da707abd9d4c/start

During handling of the above exception, another exception occurred:

...


raise cls(e, response=response, explanation=explanation)
docker.errors.APIError: 500 Server Error for http+docker://localhost/v1.30/containers/9175e58605d9cc143ee42c1e14f7f22b945f09b0fcc7f041e348da707abd9d4c/start: Internal Server Error ("oci runtime error: container_linux.go:262: starting container process caused "process_linux.go:339: container init caused \"read init-p: connection reset by peer\""
")

opened by yangyueren 3

执行json_object，结果返回字段顺序会改变，和mariadb，mysql数据库不一样

mysql> SELECT JSON_OBJECT('id', 123, 'name', 'polardb-x'); +---------------------------------------------+ | JSON_OBJECT('id', 123, 'name', 'polardb-x') | +---------------------------------------------+ | {"name":"polardb-x","id":123} | +---------------------------------------------+ 1 row in set (0.00 sec)

mariadb数据库的结果为，这应该是和mysql数据库一样 MariaDB [(none)]> SELECT JSON_OBJECT('id', 123, 'name', 'polardb-x'); +---------------------------------------------+ | JSON_OBJECT('id', 123, 'name', 'polardb-x') | +---------------------------------------------+ | {"id": 123, "name": "polardb-x"} | +---------------------------------------------+ 1 row in set (0.000 sec)

opened by co63oc 3
polardb-x支持udf
数据库中 SQL 函数可以大致分为以下几类：内部函数、系统生成的函数、用户定义的函数(UDF)，其中用户定义的函数是指由用户通过 CREATE FUNCTION 显示创建并命名的函数，其语义也由用户自己确定，能够在一定程度上扩展数据库的功能。目标效果：

PolarDB-X 支持 CREATE FUNCTION 创建 UDF

支持一些系统视图，可以方便查询和管理已注册 UDF
opened by joeCarf 3
Let the 32 databases limit configurable

Try to deploy LMS product with polardbx as it database, We need to create database over 100 schools. Then I have been told about that the limit about number of databases is 32, ant Can't adjust by config file. Wish this limit can be configurable and please tell me why we really need this limitation?

opened by datouxia 3

pxd layout failed.docker.errors.ContainerError: Command '/init' in image 'polardbx/polardbx-init:latest' returned non-zero exit status 255: b''

os info

[root@VM ~]# cat /etc/redhat-release
CentOS Linux release 7.3.1611 (Core)

[root@VM ~]# uname -a
Linux VM 3.10.0-1127.19.1.el7.x86_64 #1 SMP Tue Aug 25 17:23:54 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux

[root@VM ~]# python3 --version
Python 3.6.8

run pxd layout output

Digest: sha256:5c0b19954c18fd2347eee46dffd82c4babf7e57b2e405d8b4ccb4d5a647c19a7
Status: Downloaded newer image for docker.io/polardbx/xstore-tools:latest
Processing  [#########---------------------------]   25%    create gms node
Processing  [############------------------------]   33%    create gms db and tables
Processing  [###############---------------------]   41%    create PolarDB-X root account
Processing  [##################------------------]   50%    create dn
Processing  [#####################---------------]   58%    register dn to gms
Processing  [########################------------]   66%    create cn
--- Logging error ---
Traceback (most recent call last):
  File "/usr/local/lib/python3.6/site-packages/deployer/decorator/decorators.py", line 32, in wrapper
    ret = func(*args, **kw)
  File "/usr/local/lib/python3.6/site-packages/deployer/pxc/polardbx_cluster.py", line 380, in _create_cn_containers
    name=self.pxc_name + '-cn-init-' + ''.join(secrets.choice(string.ascii_letters) for i in range(4)))
  File "/usr/local/lib/python3.6/site-packages/docker/models/containers.py", line 841, in run
    container, exit_status, command, image, out
docker.errors.ContainerError: Command '/init' in image 'polardbx/polardbx-init:latest' returned non-zero exit status 255: b''

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/usr/local/lib/python3.6/site-packages/deployer/util/sqlite_manager.py", line 35, in execute_update
    cursor.execute(sql)
sqlite3.OperationalError: near "' in image '": syntax error

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/usr/local/lib/python3.6/logging/handlers.py", line 71, in emit
    if self.shouldRollover(record):
  File "/usr/local/lib/python3.6/logging/handlers.py", line 187, in shouldRollover
    msg = "%s\n" % self.format(record)
  File "/usr/local/lib/python3.6/logging/__init__.py", line 840, in format
    return fmt.format(record)
  File "/usr/local/lib/python3.6/logging/__init__.py", line 577, in format
    record.message = record.getMessage()
  File "/usr/local/lib/python3.6/logging/__init__.py", line 338, in getMessage
    msg = msg % self.args
TypeError: not all arguments converted during string formatting
Call stack:
  File "/usr/local/bin/pxd", line 8, in <module>
    sys.exit(main())
  File "/usr/local/lib/python3.6/site-packages/click/core.py", line 1137, in __call__
    return self.main(*args, **kwargs)
  File "/usr/local/lib/python3.6/site-packages/click/core.py", line 1062, in main
    rv = self.invoke(ctx)
  File "/usr/local/lib/python3.6/site-packages/click/core.py", line 1668, in invoke
    return _process_result(sub_ctx.command.invoke(sub_ctx))
  File "/usr/local/lib/python3.6/site-packages/click/core.py", line 1404, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "/usr/local/lib/python3.6/site-packages/click/core.py", line 763, in invoke
    return __callback(*args, **kwargs)
  File "/usr/local/lib/python3.6/site-packages/deployer/pxd.py", line 40, in tryout
    create_tryout_pxc(name, cn_replica, cn_version, dn_replica, dn_version, cdc_replica, cdc_version)
  File "/usr/local/lib/python3.6/site-packages/deployer/pxc/polardbx_manager.py", line 35, in create_tryout_pxc
    pxc.create()
  File "/usr/local/lib/python3.6/site-packages/deployer/pxc/polardbx_cluster.py", line 143, in create
    result = create_task()
  File "/usr/local/lib/python3.6/site-packages/deployer/decorator/decorators.py", line 37, in wrapper
    update_polardbx_record(pxc, ex)
  File "/usr/local/lib/python3.6/site-packages/deployer/decorator/decorators.py", line 52, in update_polardbx_record
    SQLiteManager.execute_update(sql)
  File "/usr/local/lib/python3.6/site-packages/deployer/util/sqlite_manager.py", line 38, in execute_update
    logger.error("execute update for xdb failed, xdb: %s, sql: %s", SQLiteManager.db_file, sql, e)
Message: 'execute update for xdb failed, xdb: %s, sql: %s'
Arguments: ('/root/.pxd/polardbx.db', "replace into polardbx_cluster ('id', 'gmt_created', 'gmt_modified', 'pxc_name', 'pxc_status', 'cn_replica', 'cn_version', 'dn_replica', 'dn_version', 'leader_only', 'root_account', 'root_password', 'details', 'cdc_replica', 'cdc_version') values (NULL, date('now'), date('now'), 'pxc-tryout', 'failed', '1', 'latest', '1', 'latest', 'True', 'polardbx_root', 'qdWlIjKB', 'Command '/init' in image 'polardbx/polardbx-init:latest' returned non-zero exit status 255: b''', '1', 'latest')", OperationalError('near "\' in image \'": syntax error',))
--- Logging error ---
Traceback (most recent call last):
  File "/usr/local/lib/python3.6/site-packages/deployer/decorator/decorators.py", line 32, in wrapper
    ret = func(*args, **kw)
  File "/usr/local/lib/python3.6/site-packages/deployer/pxc/polardbx_cluster.py", line 380, in _create_cn_containers
    name=self.pxc_name + '-cn-init-' + ''.join(secrets.choice(string.ascii_letters) for i in range(4)))
  File "/usr/local/lib/python3.6/site-packages/docker/models/containers.py", line 841, in run
    container, exit_status, command, image, out
docker.errors.ContainerError: Command '/init' in image 'polardbx/polardbx-init:latest' returned non-zero exit status 255: b''

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/usr/local/lib/python3.6/site-packages/deployer/util/sqlite_manager.py", line 35, in execute_update
    cursor.execute(sql)
sqlite3.OperationalError: near "' in image '": syntax error

opened by clot09 3

安装步骤疑问

源码安装参考文档中

启动PolarDB-X DN
此步骤启动一个mysql进程，作为metadb和dn
参考附录中的mysql配置文件，可进行相应修改，默认使用 4886 作为 mysql端口，32886 作为私有协议端口
默认使用 /u01/my3306 作为mysql数据目录，可以修改成其他目录
注意：启动 DN 需要使用非 root 账号完成

启动mysql：

mkdir -p /u01/my3306/{data,log,run,tmp,mysql}
/u01/mysql/bin/mysqld --defaults-file=my.cnf --initialize-insecure
/u01/mysql/bin/mysqld --defaults-file=my.cnf

这里的mysql是需要单独装一个mysql server端？还是客户端？还是需要软链接到对应目录下？

question

opened by Hoxxx 3

Fix a 'recreate compound primary key' issue

Problems solved in this PR

Issue #114

Changes proposed in this PR

Also hide metadata for dropped primary key before executing physical DDL operations.

Tests

Added a new test case: AlterTableTest.testAlterTableDropAddCompPrimaryKey()

opened by mybjlife 1
close #101 fix rename table failed
Problems solved in this PR

Fix the bug that failed to rename table.

Changes proposed in this PR

rename table

Tests

[ ] Unit test

[ ] DML_DDL test

Release note

None
opened by hustStt 1
Fix (issuse: #104) : Cast error when 'set password'
Problems solved in this PR

issuse: https://github.com/ApsaraDB/galaxysql/issues/104

Changes proposed in this PR

Tests

[ ] Unit test

[ ] DML_DDL test

Release note

None
opened by joeCarf 0

mysql client occasionally encounters errors

mysql> show create table t1;
+-------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| TABLE | CREATE TABLE                                                                                                                                                                                                                                                        |
+-------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| t1    | CREATE TABLE `t1` (
        `id` bigint(11) NOT NULL AUTO_INCREMENT,
        `score` tinyint(4) NOT NULL,
        `grade` tinyint(4) NOT NULL,
        PRIMARY KEY (`id`),
        KEY `ig` (`grade`)
) ENGINE = InnoDB AUTO_INCREMENT = 12 DEFAULT CHARSET = utf8
PARTITION BY KEY(`id`)
PARTITIONS 8 |
+-------+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1 row in set (0.00 sec)

mysql> alter table t1 add age tinyint not null;
ERROR 2013 (HY000): Lost connection to MySQL server during query
No connection. Trying to reconnect...
ERROR 2013 (HY000): Lost connection to MySQL server at 'reading initial communication packet', system error: 0
ERROR:
Can't connect to the server

mysql> alter table t1 add age tinyint not null;
No connection. Trying to reconnect...
Connection id:    3
Current database: d3

Query OK, 0 rows affected (3.63 sec)

opened by yejr 1

Releases(galaxysql-5.4.15-16669515)

galaxysql-5.4.15-16669515(Oct 28, 2022)

RELEASE NOTE

Feature enhancement

Add New Sequence: a globally unique, monotonically increasing, high-performance sequence. Support for specifying and modifying locality properties in AUTO mode. Support partition-level heatmap. Support for automatic partition splitting. Support stored procedure. Suuport user defined function. Support partitioned table change at table-level. support insert overwrite. Support online modification of column types. Support server-side cursor. Add sharding advise syntax to recommend table splitting method. Support archive history of DDL task. Support STATEMENTS_SUMMARY and STATEMENTS_SUMMARY_HISTORY table. Add ALTER TABLE PARTITIONS syntax to modify the table's number of partitions. Support automatically pause scheduled tasks of TTL-table. Support configurate and detecte password complexity and expiration date. Add audit user's login and logout (disabled by default). Optimize read and write performance of auto-mode table. Optimize memory usage of the Scan operator during streaming mergesort. Optimize type comparison logic of new partition table. Optimize the logic of set global variables. Optimize the speed of data migration when creating global secondary indexes, scaling and shrinking, changing partitions, and so on. Optimize the execution of insert, update, delete, and select statements. Optimize memory usage of query with in clause. Optimize logic of permission verification. Optimize the logic of change partitions on auto-partitioned tables. Remove the restriction that user tables cannot have the same name as system tables. Optimize the implementation of optimize table.

Bugfix

Fixed an issue where read-only traffic could not be switched back to the primary node in read/write splitting scenarios when the log-copy thread is interrupted. Fixed an issue that the alias of having clause in physical SQL is incorrect. Fixed an issue that SQL throttling might cause client blocking. Fixed an compatibility issue of SQL throttling when table name with backticks. Fixed the compatibility issue when cachePrepStmts and useServerPrepStmts were enabled at the same time and the mysql connector is 8.0. Fixed the broadcast shuffle selection issue in Outer Join in MPP mode. Fixed an incorrect result issue of union distinct query in MPP mode. Fix an icompatibility issue of primary key for Navicat. Fix an error of 'IN' statement in Prepare mode. Fix that the value of 'Current timestamp' don't change when used in the same session. Fix an error of 'Update' statement for BINARY type. Fix an incurrent result issue of JSON operator '->'. Fix an incorrect result issue of to_base64 function when parsing binary variables Fix an array out of bounds issue of HashGroupJoin operator. Fix an incorrect result issue of AES encryption function. Fix an incorrect result issue of covering index. Fix an incorrect result issue of window functions in some cases. Fix that the table name is not correctly replaced in single table Update statement with subquery. Fix a NPE issue in multi table Update command. Fix an occasional error in subquery. Fix an incorrect result issue of yyyyweek in a range query which cross years. Fix that the actual index selected is different from 'Explain Execute' in some case. Fix an unexpected pushdown issue of non pushdown functions in some scenarios. Fix an error when trying to create an user without host. Fix an error of escape character of JSON_QUOTE function. Fix a NPE issue of 'Explain Optimizer'. Fix a lock wait timeout issue when recording physical DDL status during DDL task execution. Fix an empty or incomplete result issue of 'Trace ddl' Fix an incompatibility issue of 'show rule' in auto partition database. Fix an index metadata lost issue when adding columns and adding indexes at the same time. Fix a statistics lost issue of 'Rename Table' or 'Change Columns'. Fix a plan pollution issue when the comparative object inside changed. Fix a primary key lost issue of global unique index in auto partition database. Fix that GSI related indexes are pushed down incorrectly while using 'ignore index' in auto partition database. Fix that single table cannot be pushed down in the auto partition database. Fix a NPE issue when initializing sequence in the read-only instance. Fix an error when changing the column name to a special character. Fix an error of 'alter table modify column after'. Fix an error when using constant expressions as partition keys. Fix an error of 'DROP CLUSTERED GSI'. Fix an incorrect result issue of month partition function in a range query. Fix an error of 'Truncate Table' with LOCAL FULLTEXT KEY Fix a concurrency security issue when unwarping subqueries. Fix an empty result issue of 'show status' Fix an incorrect result issue of when join broadcast table and single table in auto partition database. Fix that read/write separation does not take effect in read-only transactions. Fix an error of rebalance task when trying to update system tables. Fix a NPE issue during MDL lock acquisition. Fix that table name maybe wrong in CDC Fix an error of executing DDL with CDC.

发布说明

特性更新

新增支持全局唯一、单调递增、高性能的New Sequence。新增支持在AUTO模式下指定和修改locality属性。新增分区热力图功能。新增支持分区自动分裂。新增支持存储过程。新增支持自定义函数。新增支持表级别的分区表变更。新增支持Insert Overwrite语法。新增支持在线修改列类型。新增支持服务端的游标。新增sharding advise语法进行表的拆分方式推荐。新增DDL任务历史数据归档。新增支持STATEMENTS_SUMMARY和STATEMENTS_SUMMARY_HISTORY元数据库表。新增支持ALTER TABLE PARTITIONS语法，一键修改分区数。新增支持自动暂停TTL表的定时任务。新增支持密码复杂度、有效期的配置和检测。新增支持用户的登入登出审计功能（默认关闭）。优化AUTO库的读写并行度及其读写性能。优化Scan算子在流式归并排序过程中的内存使用。优化新分区表DML类型比较逻辑。优化set global变量的设置逻辑。优化创建全局二级索引、扩缩容、分区变更等操作时数据迁移的速度。优化insert、update、delete和select语句的执行性能。优化IN查询内存占用过高的问题。优化权限校验流程。优化默认自动分区的表进行分区变更的逻辑。解除用户表不能与系统表同名的限制。优化Optimize table的实现。

问题修复

修复当日志复制线程中断时，读写分离场景中只读流量无法切回主库的问题。修复物理SQL的Having别名语法不正确的问题。修复SQL限流可能导致客户端阻塞的问题。修复SQL限流匹配表名时带反引号，无法匹配表名的问题。修复8.0驱动下同时开启cachePrepStmts和useServerPrepStmts时的兼容性问题。修复MPP执行模式下Outer Join的Broadcast Shuffle选择问题。修复union distinct查询在MPP执行模式下查询结果不正确的问题。修复PolarDB-X对于Navicat主键元信息的兼容性。修复Prepare模式下，IN语句转子查询报错的问题。修复同一个会话中使用Current timestamp时，值不会变化的问题。修复执行UPDATE语句时BINARY类型处理不正确的问题。修复JSON操作符->在表达式中解析不正确的问题。修复to_base64解析binary变量不正确的问题。修复HashGroupJoin算子存在数组越界的问题。修复AES加密函数返回结果不符合预期的问题。修复使用覆盖索引查询时返回错误结果的问题。修复某些情况下窗口函数计算错误的问题。修复单表Update语句时子查询的表名没有被正确替换的问题。修复Update命令中的多表NPE的问题。修复部分子查询偶发报错的问题。修复对于yyyyweek跨年周闭区间范围查询分片缺失的问题。修复私有协议传递执行计划时，Explain Eexcute结果与实际选择索引不同的问题。修复不可下推函数在某些场景下的非预期下推问题。修复没有带host的用户创建语句报错的问题。修复非下推JSON_QUOTE函数的转义字符问题。修复Explain Optimizer抛空指针的问题。修复DDL任务执行过程中，记录物理DDL状态时可能发生的Lock wait timeout问题。修复trace ddl后，可能出现show ddl为空或者不完整的问题。修复SHOW RULE命令不兼容新分区表的问题。修复加列同时创建索引后缺失索引元数据的问题。修复Rename Table或Alter Table变更列后统计信息被清除的问题。修复Comparative对象变更导致计划污染的问题。修复AUTO库下全局唯一索引丢失主键信息的问题。修复AUTO模式数据库中ignore index时GSI相关的索引被错误下推的问题。修复AUTO模式数据库中单表无法下推的问题。修复只读实例中Sequence初始化失败上报NPE的问题。修复ALTER TABLE将列名改为特殊字符时，元数据处理失败的问题。修复alter table modify after失败的问题。修复使用常量表达式作为分区键时，缺少语法校验并导致建表失败的问题。修复执行DROP CLUSTERED GSI失败问题。修复使用month分区函数时区间范围查询结果不正确的问题。修复带LOCAL FULLTEXT KEY的表无法truncate的问题。修复子查询消除时并发安全的问题。修复show status视图返回空的问题。修复AUTO模式数据库下广播表与单表Join下推时，选错物理库导致执行报错的问题。修复读写分离在只读事务下未生效的问题。修复rebalance任务更新系统表的SQL错误的问题。修复MDL锁获取过程中抛空指针的问题。修复CDC打标信息中表名不一致的问题。修复CDC启动时执行DDL失败的问题。
Source code(tar.gz)
Source code(zip)
galaxysql-5.4.15-alpha-1(Oct 28, 2022)

Source code(tar.gz)
Source code(zip)
galaxysql-5.4.13-16615127(Aug 26, 2022)

RELEASE NOTE

Feature enhancement

support dragonwell11 #66 support bool type #76 support add channel for change replication filter support add dest applier type for change master support CI/CD optimize shutdown.sh script optimize use metaDb user rather than metaDb root user when initialization optimize dockfile, use galaxybaseimage optimize change protoc version to 3.17.3 optimize doc

Bugfix

fix overflow when calc direct memory size. fix ccl reschedule return error when exception is threw fix NEVarcharColCharConstVectorizedExpression

发布说明

特性更新

支持 dragonwell11 #66 支持 bool 类型 #76 支持修改单个主备链路的配置支持修改主备链路的写入方式支持 CI/CD 流程优化 shutdown.sh 脚本优化配置初始化步骤，使用 metadb_user 而非 metadb_root_user 连接元数据库优化 dockerfile, 使用 galaxybaseimage 作为基础镜像优化 protoc 版本到 3.17.3 优化帮助文档

问题修复

修复私有协议预分配内存在系统内存过大时，数值计算溢出的问题修复被SQL限流设置为等待的查询偶发无法返回的问题修复 NEVarcharColCharConstVectorizedExpression 错误
Source code(tar.gz)
Source code(zip)
galaxysql-5.4.13-16615127-alpha-1(Aug 26, 2022)

Source code(tar.gz)
Source code(zip)
galaxysql-5.4.13-16512132(Apr 29, 2022)
RELEASE NOTE

Feature enhancement

support file storage quick-start

support CREATE FileStorage {'OSS'| 'LOCAL_DISK' } with (xxx)

support DROP FileStorage {'OSS'| 'LOCAL_DISK' }

support CREATE TABLE t_oss LIKE t_innodb engine = {'OSS'| 'LOCAL_DISK' } archive_mode = {'empty' | 'loading' | 'ttl'}

support ALTER TABLE new_table EXCHANGE PARTITION partition_name WITH TABLE old_table;

support SELECT * from table AS OF TIMESTAMP 'yyyy-mm-dd hh:mm;dd'

support ALTER TABLE table_name DROP FILE file_name [,file_name...]

support SHOW FileStorage

support SHOW FILES FROM oss_table

support SHOW CACHE_STATS

support SHOW ARCHIVE

support UNARCHIVE table source_table

发布说明

特性更新

支持 file storage quick-start

支持 CREATE FileStorage {'OSS'| 'LOCAL_DISK' } with (xxx)

支持 DROP FileStorage {'OSS'| 'LOCAL_DISK' }

支持 CREATE TABLE t_oss LIKE t_innodb engine = {'OSS'| 'LOCAL_DISK' } archive_mode = {'empty' | 'loading' | 'ttl'}

支持 ALTER TABLE new_table EXCHANGE PARTITION partition_name WITH TABLE old_table;

支持 SELECT * from table AS OF TIMESTAMP 'yyyy-mm-dd hh:mm;dd'

支持 ALTER TABLE table_name DROP FILE file_name [,file_name...]

支持 SHOW FileStorage

支持 SHOW FILES FROM oss_table

支持 SHOW CACHE_STATS

支持 SHOW ARCHIVE

支持 UNARCHIVE table source_table

Source code(tar.gz)
Source code(zip)
galaxysql-5.4.13-16512132-alpha-2(Apr 29, 2022)

Source code(tar.gz)
Source code(zip)
galaxysql-5.4.13-16511320-alpha-1(Apr 28, 2022)

Source code(tar.gz)
Source code(zip)
galaxysql-5.4.13(Mar 31, 2022)
RELEASE NOTE

Feature enhancement

Support create database with auto mode (two optional mode: auto or drds mode) , drds is the default option

Support create table with MySQL compatible syntax for partition-table， including hash/key/list/range types, but not support subpartition yet.

Support dynamic partition pruning, including collapse conditions of constant/literal， prefix match prunning, combine overlapping range scan etc.

Support join-wise push down for partitioning table.

Support add/drop/split/merge/move partition operation for partitioning table.

Support create GSI for hash/key/list/range partitioning table.

Support use primary key as partition key and compute the partition member automatically for the tables in auto-mode database.

Support to change the partition strategy for tables in auto-mode database.

Support to add primary key as additional partition key for GSI to avoid the GSI table data skew.

Support Scale-in.

Support TTL in auto-mode database.

Optimize check table to validate the meta consistency(including partitioning information, column information) for table and its GSI.

Support broadcast table as candiate for SQL Advisor feature.

Support instant add column.

Support to collect all information for optimizer when use explain statistics command.

Support limit the search space of CBO and reduce the optimization time of complex queries.

Optimize the validation data phase of ddl operation to speed up the GSI/Scale in/out operation.

Support MySQL compatible replica command.

Support GalaxyEngine cluster.

Bugfix

Fix the NPE and sort error issues of heuristic join reorder

Fix the RelNode rowtype error cause by AccessPathRule

Fix the GSI data inconsistency issue when the collaction is utf8mb4_0900_ai_ci

Fix the maximum login times cannot take effect

Fix inaccurate estimate size in some block data types

Fix missing accuracy of Deciaml for GalaxySQL communication protocol.

Fix the Byte and Char types error when push down execution plan to MySQL in GalaxySQL communition protocol.

Fix index metadata is missing after alter table add/drop primary Key

Fix the shards/partitions definition inconsistency issue caused by interrupt or timeout during alter Table

Fix can't rollback automatically when all the sub-operations are failed for alter table

Fix possible NPE when input the empty resultset for the sortAgg operation

发布说明

特性更新

新增支持创建数据库指定建表模式(新的分区表模式与老的分库分表模式)，默认是分库分表模式

新增支持使用 MySQL分区表语法创建一级分区的分区表，分区策略包括Hash/Range/List等

新增支持分区表的动态裁剪能力，包括支持分区列条件的常量折叠、区间合并以及前缀查询裁剪等功能

新增支持分区表的JOIN计算下推

新增提供分区表的分区管理能力，包括分区的添加、删除、分裂、合并与迁移等功能

新增提供表组及其他能力（包括表组的创建、删除、变更等），支持分区变更期间JOIN计算下推不受影响

新增支持全局索引表使用MySQL分区表语法并按Hash/Range/List等分区策略进行分区

新增自动拆分支持使用分区表语法

新增拆分变更增加支持分区表

新增新分区表GSI自动拆分会携带主键，可以处理GSI热点问题

新增支持实例的缩容

新增支持分区表的TTL及其管理能力（包括调整TTL的初始时间与时间间隔等）

优化 Check Table 指令，支持校验主表分区、索引表分区与列定义等元数据一致性

新增 SQL Advisor支持推荐广播表

新增支持Instant Add Column功能

新增支持Explain Statistics拉取优化器优化需要的所有信息

新增限制cbo的搜索空间，减少复杂查询的优化耗时

优化部分DDL后台操作的数据校验任务的性能，使GSI/扩缩容DDL变更操作加速

新增支持兼容MySQL的Replica相关指令

新增支持存储节点PAXOS三节点集群

问题修复

修复 heuristic join reorder的空指针和排序报错的问题

修复 AccessPathRule导致的RelNode rowtype错误的问题

修复当 collate 为 utf8mb4_0900_ai_ci 时，建 GSI 导致数据不一致的问题。

修复最大登录次数无法生效的问题

修复某些block中estimateSize不准确的问题

修复私有协议下Deciaml精度缺失是问题

修复私有协议下执行计划传输byte和char类型对比报错的问题

修复 Alter Table Add/Drop Primary Key后索引元数据缺失问题

修复 Alter Table期间外部中断或物理分片超时导致的表结构不一致问题

修复 Drop index所有物理分片都失败时不自动回滚的问题

修复 sortAgg在结果集为空时可能抛NPE的问题

Source code(tar.gz)
Source code(zip)
galaxysql-5.4.13-alpha-1(Mar 30, 2022)

Source code(tar.gz)
Source code(zip)
galaxysql-5.4.12(Jan 28, 2022)
RELEASE NOTE

Feature enhancement

Support access GalaxySQL from maxwell and debezium

Support the separation of reading and writing for RW node

Support set TLS version based on jdk version

Support statement SHOW GRANTS FOR CURRENT_USER

Support disable cost estimation for LogicalView with hint ENABLE_LOGICALVIEW_COST=false

Support auto rollback for ALTER TABLE ddl task

Add precheck for ALTER TABLE task

Add records for information_shema in result of SHOW DATASOURCES

Avoid pushing uncorrelate node into lv with variables set

Try transform filter downside of projects with uncorrelate scalar subquery

Change the level of spm null point log from error to info

More literal expression supported for SET statement

Optimize the execution of apply executor for IN subquery

Optimize duplicate check for insert statement executed with logical multi write

Optimize physical query with limit

Remove PostPlanner segmented check

Optimize Compare function for FactorCostComparator

Support huge memory size in startup.sh

Limit length of SHOW JOBS result to 500

Add sample rate to accurate count num from TopN

Rollback transaction after ERR_TRANS_DEADLOCK exception

Bugfix

Fix AES encryption when key length greater than 16 bytes

Fix the issue of LIMIT + physical sql cache leads to empty result set of SELECT statement with LIMIT clause

Fix alter table add index without specifying index name issue that can cause failure of inserting index metadata

Fix index meta missed issue when altering table add multiple indexes with the same columns

Fix a NPE issue of simple sequence validation

Fix DDL task issue of executed by 2 CN during leader election

Fix a concurrent physical ddl issue that cannot recover ddl task after task paused

Fix the issue of traceid of transaction changed after execution of ddl

Fix the issue of subtask not exit correctly while there occurs an out of memory exception

Fix the issue of MOVE DATABASE task cannot be recovered for some case

Fix cn/dn mapping maintaining issue that can cause failure of scale out task

Fix the invalidation of baseline caused by DDL

Fix the NPE issue for baseline sync

Fix the issue of plan leaking in baseline

Fix backfillRows of show stats

Fix memorypool leaking issue for some case

Fix case sensitive issue for table meta initialization that can cause some column mata not loaded

Fix virtual_statistic result not correct issue which is caused by lack of table meta

Fix physical sql for statement using inventory hint

Fix execution error of statement using SCAN HINT, which is caused by table name not surrounded by backquote

Fix execution error of BkaJoin, which is caused by physical sql like ' xx in row(EMPTY)' generated

Fix wrong physical sql of order by [index]

Fix sql template cache when there exists scalar subquery in logicalview.

Fix rex literal transform issue in subquery apply executor.

Fix for get cost after execution

Fix npe issue when optimizer trying to get the statisic of index

Fix estimated size in some block

Fix the issue of "Snapshot Too Old" for long-lasting transaction

Fix the wrong UPDATE result issue for some case of using ON UPDATE TIMESTAMP column as partition key

Fix the issue of UPDATE column with property AUTO_INCREMENT

Fix UPSERT BIT type column issue that can cause UPSERT failure

Fix UPDATE blob type column issue that in some case can cause wrong update result

Fix setting value for AUTO_INCREMENT column issue for INSERT SELECT statement without target column

Fix logical multi-write issue for REPLACE statement in isolation level read committed

Fix ddl validation for not support using SET type column as sharding key

Fix undefined user variable for sum

Fix the issue of SetHandler#parserValue

Fix DESC issue that can cause executeion failure on system table

Fix execution issue of statament EXPLAIN ANALYZE LIMIT

Fix exception "Failed to create vectorized expression"

Fix aggregation function execution issue for UNION statement without actual table

Fix the issue of SHOW DS STATUS and KILL ALL not work at RO instance in some case

Fix execution issue of SHOW TABLES using readonly connection

Fix the issue that condition derivation not adapted for subquery materialize

Fix a issue about login

Fix issues about MySQL compatibility

发布说明

特性更新

新增 ALTER TABLE 校验，提前拒绝执行不支持的 ALTER TABLE 语句

新增 CN 启动参数，支持设置 LargePage

新增 SHOW DATASOURCES 中加入 information_schema 的相关信息

新增兼容 debezium 使用到的特殊 sql 语句

新增在主实例上引入读写分离能力

新增基于 JDK 版本自适应设置 TLS 版本号

新增支持 SHOW GRANTS FOR CURRENT_USER 语句

新增通过 HINT ENABLE_LOGICALVIEW_COST 关闭 LogicalView 的代价估算

新增限制 SHOW JOBS 的 phyProcess 最长为500

优化非相关联子查询物理 SQL 结构

优化 IN 类 APPLY 子查询的执行效率, APPLY 执行器支持物化方式执行

优化逻辑多写 Duplicate Check 的执行性能

优化 SET 语句支持的表达式范围

优化 ALTER TABLE 支持自动回滚

优化物理 SQL 包含 LIMIT 的场景

优化去除 PostPlanner segmented check

优化禁用集合类型作为拆分键

优化 FactorCostComparator 中的 Compare 函数性能

问题修复

修复 AES加密函数当密钥长度超过16字节时结果不正确

修复 LIMIT + 物理 SQL CACHE 导致 LIMIT 查询结果为空的问题

修复 ALTER TABLE 为同一列多次新增不带索引名的索引时元数据插入失败的问题

修复 ALTER TABLE 新增多个包含同一字段的索引后，部分索引元数据缺失

修复 ALTER TABLE 语句通过 Auto_Increment 子句修改 Sequence 起始值时发生 NPE

修复 CN 发生切主时，2 个 CN 可能同时执行同一个 DDL 任务的问题

修复 DDL 分库间并行策略下的任务中断后，任务恢复执行卡住的问题

修复 DDL 导致事务 ID 发生变化

修复 OOM 时 DDL 引擎中子线程没有正确退出的问题

修复 MOVE DATABASE 任务中断恢复的问题

修复创建建连接池异常，导致 scaleout 任务中断的问题

修复 SPM 在 fix sql 时不会受 DDL 影响而失效

修复 baseline fix 时会概率性 sync 失败的问题

修复在复杂 SQL 较多的场景中, baseline 存储的 plan 有概率超出最大值泄露的问题

修复 TopN 因采集数据过大导致的估算问题

修复 SHOW STATS 中的 backfillRows 统计

修复异常情况下，memory pool 对象未被 destory 情况

修复批量加载 table meta 时候，表名的大小写会导致列加载不到的问题

修复 virtual_statistic 视图因 table meta 获取不到导致的展示问题

修复热点更新的下推 SQL

修复通过 HINT 下推的 SQL 表名没有带反引号，导致报错的问题

修复物理 SQL 中出现' xx in row(EMPTY)' 导致报错的问题

修复物理 SQL 生成 ORDER BY 带 index 的问题

修复物理 SQL CACHE 涉及子查询问题

修复执行结束后 getCost 报错问题

修复统计信息涉及索引部分场景下 NPE 问题

修复部分 block 的 estimatedSize 不准确问题

修复部分场景下，只读实例上 SHOW DS STATUS 和 KILL ALL 指令不能正确执行的问题

修复长事务导致 Snapshot too old 报错的问题

修复死锁检测报错 ERR_TRANS_DEADLOCK 之后未主动回滚事务

修复 ON UPDATE TIMESTAMP 作为拆分键时，部分场景下 UPDATE 结果不符合预期的问题

修复 UPDATE 自增列的问题

修复 UPSERT BIT 类型报错的问题

修复部分场景下 UPDATE blob 字段结果不符合预期的问题

修复使用隐式目标列 + LIMIT 且不可下推的 INSERT SELECT 语句，主键没有使用 Sequence 填充的问题

修复禁止 LogicalReplace 在隔离级别为 RC 时的下推

修复未知的用户变量问题

修复 SetHandler#parserValue 的问题

修复 DESC 系统表报错的问题

修复 EXPLAIN ANALYZE + LIMIT 报错的问题

修复 Failed to create vectorized expression 错误

修复不包含表的 FROM UNION 上聚合函数执行报错的问题

修复使用只读连接串执行 SHOW TABLES 报错问题

修复子查询物化适配条件推导

修复登陆相关问题

修复 SQL 兼容性相关问题修复

Source code(tar.gz)
Source code(zip)
galaxysql-5.4.12-alpha-1(Jan 27, 2022)

Source code(tar.gz)
Source code(zip)