dsql

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Amazon Aurora DSQL Skill

Amazon Aurora DSQL 技能文档

Aurora DSQL is a serverless, PostgreSQL-compatible distributed SQL database. This skill covers direct query execution via MCP tools, schema management, migrations, multi-tenant isolation, IAM auth, and bulk data loading via

aurora-dsql-loader

Aurora DSQL是一款兼容PostgreSQL的无服务器分布式SQL数据库。本技能文档涵盖通过MCP工具执行直接查询、模式管理、迁移、多租户隔离、IAM认证，以及通过

aurora-dsql-loader

进行批量数据加载。

Reference Files

参考文档

Load these files as needed for detailed guidance:

根据需要加载以下文件获取详细指导：

Core:

核心类：

Reference	When to Load	Contains
development-guide.md	ALWAYS before schema changes or DB operations	Best practices, DDL rules, transaction limits, app-layer referential integrity
language.md	MUST load for language-specific choices	Driver selection, DSQL Connectors, connection code
access-control.md	MUST load for roles, grants, or sensitive data	Scoped role setup, IAM-to-database role mapping
troubleshooting.md	SHOULD load for errors or unexpected behavior	OCC errors, connection failures, cluster state errors, token expiry, DDL rejection causes
dsql-examples.md	Load for implementation examples	Multi-tenant schema examples, batch operations, FK validation patterns, connection pooling
onboarding.md	User requests "Get started with DSQL"	Interactive step-by-step guide
occ-retry-patterns.md	MUST load for OCC retry code or conflict mitigation	DSQL Connectors, manual retry pattern, idempotent design

参考文档	加载时机	包含内容
development-guide.md	进行模式变更或数据库操作前必须加载	最佳实践、DDL规则、事务限制、应用层参照完整性
language.md	进行语言相关选型时必须加载	驱动选择、DSQL连接器、连接代码
access-control.md	处理角色、权限或敏感数据时必须加载	限定范围角色配置、IAM到数据库角色的映射
troubleshooting.md	遇到错误或异常行为时建议加载	OCC错误、连接失败、集群状态错误、令牌过期、DDL拒绝原因
dsql-examples.md	需要实现示例时加载	多租户模式示例、批量操作、FK验证模式、连接池
onboarding.md	用户请求“开始使用DSQL”时加载	交互式分步指南
occ-retry-patterns.md	需要OCC重试代码或冲突缓解方案时必须加载	DSQL连接器、手动重试模式、幂等设计

MCP:

MCP类：

Reference	When to Load	Contains
mcp-setup.md	Always for MCP server guidance	Setup instructions, 2 configuration options
mcp-tools.md	For MCP tool syntax and examples	Tool parameters, input validation
dsql-lint.md	MUST load before running `dsql_lint` or processing external SQL	Tool reference, fix statuses, unfixable error resolution

参考文档	加载时机	包含内容
mcp-setup.md	进行MCP服务器指导时始终加载	配置说明、2种配置选项
mcp-tools.md	需要MCP工具语法及示例时加载	工具参数、输入验证
dsql-lint.md	运行 `dsql_lint` 或处理外部SQL前必须加载	工具参考、修复状态、无法修复错误的解决方案

DDL Migrations:

DDL迁移类：

Reference	When to Load	Contains
ddl-migrations/overview.md	MUST load for DROP COLUMN, ALTER TYPE, DROP CONSTRAINT	Table recreation pattern, verify & swap
ddl-migrations/column-operations.md	DROP COLUMN, ALTER TYPE, SET/DROP NOT NULL/DEFAULT	Column-level migration patterns
ddl-migrations/constraint-operations.md	ADD/DROP CONSTRAINT, MODIFY PRIMARY KEY	Constraint and structural changes
ddl-migrations/batched-migration.md	Tables exceeding 3,000 rows	Batching patterns, progress tracking

参考文档	加载时机	包含内容
ddl-migrations/overview.md	执行DROP COLUMN、ALTER TYPE、DROP CONSTRAINT时必须加载	表重建模式、验证与替换
ddl-migrations/column-operations.md	执行DROP COLUMN、ALTER TYPE、SET/DROP NOT NULL/DEFAULT时	列级迁移模式
ddl-migrations/constraint-operations.md	执行ADD/DROP CONSTRAINT、MODIFY PRIMARY KEY时	约束与结构变更
ddl-migrations/batched-migration.md	处理超过3000行的表时	批处理模式、进度跟踪

MySQL Migrations:

MySQL迁移类：

Reference	When to Load	Contains
mysql-migrations/type-mapping.md	MUST load for MySQL → DSQL migration	Data type mappings, feature alternatives
mysql-migrations/ddl-operations.md	Translating MySQL DDL to DSQL	AUTO_INCREMENT, ENUM, SET, FK patterns
mysql-migrations/full-example.md	Complete MySQL table migration	End-to-end example with decision summary

参考文档	加载时机	包含内容
mysql-migrations/type-mapping.md	进行MySQL → DSQL迁移时必须加载	数据类型映射、特性替代方案
mysql-migrations/ddl-operations.md	将MySQL DSQL转换为DSQL时	AUTO_INCREMENT、ENUM、SET、FK模式
mysql-migrations/full-example.md	进行完整MySQL表迁移时	端到端示例及决策总结

PostgreSQL Migrations:

PostgreSQL迁移类：

Reference	When to Load	Contains
pg-migrations/type-mapping.md	MUST load for PG → DSQL type questions	C collation rules, NUMERIC precision, JSON/JSONB
pg-migrations/fk-replacement.md	MUST load for FK validation code generation	Tenant-scoped validate_fk_*() template, cascade
pg-migrations/index-conversion.md	MUST load for unfixable index diagnostics	GIN/GiST/BRIN → btree, partial, expression indexes
pg-migrations/schema-objects.md	MUST load for ENUM, materialized views, extensions, multi-schema	ENUM → CHECK, views, role/IAM mapping
pg-migrations/multi-region.md	Multi-region, active-active, or HA questions	Architecture, geographic partitioning

参考文档	加载时机	包含内容
pg-migrations/type-mapping.md	遇到PG → DSQL类型相关问题时必须加载	C排序规则、NUMERIC精度、JSON/JSONB
pg-migrations/fk-replacement.md	需要生成FK验证代码时必须加载	租户范围的validate_fk_*()模板、级联
pg-migrations/index-conversion.md	遇到无法修复的索引诊断问题时必须加载	GIN/GiST/BRIN → btree、部分索引、表达式索引
pg-migrations/schema-objects.md	处理ENUM、物化视图、扩展、多模式时必须加载	ENUM → CHECK、视图、角色/IAM映射
pg-migrations/multi-region.md	涉及多区域、双活或高可用问题时	架构设计、地理分区

ORM Guides:

ORM指南类：

Reference	When to Load	Contains
orm-guides/overview.md	Migrating any ORM to DSQL	Adapter names, key gotchas for Django/Hibernate/Rails/SQLAlchemy

参考文档	加载时机	包含内容
orm-guides/overview.md	将任意ORM迁移到DSQL时	适配器名称、Django/Hibernate/Rails/SQLAlchemy的关键注意事项

Data Loading:

数据加载类：

Reference	When to Load	Contains
data-loading.md	Planning or running bulk loads with `aurora-dsql-loader`	Fresh-vs-warm partitions, resume/retry, `--on-conflict` semantics, throughput diagnostics

参考文档	加载时机	包含内容
data-loading.md	使用 `aurora-dsql-loader` 规划或执行批量加载时	冷/热分区、恢复/重试、 `--on-conflict` 语义、吞吐量诊断

Query Plan Explainability:

查询计划可解释性：

Reference	When to Load	Contains
query-plan/plan-interpretation.md	MUST load at Workflow 9 Phase 0	DSQL node types, Node Duration math, estimation-error bands
query-plan/catalog-queries.md	MUST load at Workflow 9 Phase 0	`pg_class` / `pg_stats` / `pg_indexes` SQL, correlated-predicate verification
query-plan/guc-experiments.md	MUST load at Workflow 9 Phase 0	GUC experiment procedures, 30-second skip protocol
query-plan/report-format.md	MUST load at Workflow 9 Phase 0	Required report structure, element checklist, support request template

参考文档	加载时机	包含内容
query-plan/plan-interpretation.md	工作流9阶段0时必须加载	DSQL节点类型、节点时长计算、估算误差范围
query-plan/catalog-queries.md	工作流9阶段0时必须加载	`pg_class` / `pg_stats` / `pg_indexes` SQL、关联谓词验证
query-plan/guc-experiments.md	工作流9阶段0时必须加载	GUC实验流程、30秒跳过协议
query-plan/report-format.md	工作流9阶段0时必须加载	要求的报告结构、元素检查清单、支持请求模板

MCP Tools Available

可用的MCP工具

The

aurora-dsql

MCP server provides these tools:

Database Operations:

readonly_query - Execute SELECT queries (returns list of dicts)
transact - Execute DDL/DML statements in transaction (takes list of SQL statements)
get_schema - Get table structure for a specific table

SQL Validation:

dsql_lint - Validate SQL for DSQL compatibility and optionally auto-fix issues. Use before executing externally-sourced SQL.

Documentation & Knowledge:

dsql_search_documentation - Search Aurora DSQL documentation
dsql_read_documentation - Read specific documentation pages
dsql_recommend - Get DSQL best practice recommendations

Note: There is no

list_tables

tool. Use

readonly_query

with information_schema.

See mcp-setup.md for detailed setup instructions. See mcp-tools.md for detailed usage and examples.

aurora-dsql

MCP服务器提供以下工具：

数据库操作：

readonly_query - 执行SELECT查询（返回字典列表）
transact - 在事务中执行DDL/DML语句（接收SQL语句列表）
get_schema - 获取指定表的结构

SQL验证：

dsql_lint - 验证SQL的DSQL兼容性并可选自动修复问题。执行外部来源的SQL前请使用该工具。

文档与知识：

dsql_search_documentation - 搜索Aurora DSQL文档
dsql_read_documentation - 读取特定文档页面
dsql_recommend - 获取DSQL最佳实践建议

注意： 没有

list_tables

工具。请使用

readonly_query

查询information_schema。

详细配置说明请查看mcp-setup.md。详细用法及示例请查看mcp-tools.md。

AWS Knowledge MCP (

awsknowledge

)

AWS知识MCP（

awsknowledge

）

Consult for verifying DSQL service limits before advising users. The numeric limits below are defaults that may change — when a user's decision depends on an exact limit, verify it first:

Limit	Default	Verify query
Max rows per transaction	3,000	`aurora dsql transaction limits`
Max data size per transaction	10 MiB	`aurora dsql transaction limits`
Max transaction duration	5 minutes	`aurora dsql transaction limits`
Max connections per cluster	10,000	`aurora dsql connection limits`
Auth token expiry	15 minutes	`aurora dsql authentication token`
Max connection duration	60 minutes	`aurora dsql connection limits`
Max indexes per table	24	`aurora dsql index limits`
Max columns per index	8	`aurora dsql index limits`
IDENTITY/SEQUENCE CACHE values	1 or >= 65536	`aurora dsql sequence cache`
Supported column data types	See docs	`aurora dsql supported data types`

When to verify: Before recommending batch sizes, connection pool settings, or schema designs where hitting a limit would cause failures; any time the exact number can affect user decision.

Fallback: If

awsknowledge

is unavailable, use the defaults above and flag that limits should be verified against DSQL documentation.

在为用户提供建议前，可通过该工具验证DSQL服务限制。以下数值限制为默认值，可能会变更——当用户的决策依赖精确限制时，请先进行验证：

限制项	默认值	验证查询语句
单事务最大行数	3,000	`aurora dsql transaction limits`
单事务最大数据量	10 MiB	`aurora dsql transaction limits`
最大事务时长	5分钟	`aurora dsql transaction limits`
单集群最大连接数	10,000	`aurora dsql connection limits`
认证令牌过期时间	15分钟	`aurora dsql authentication token`
最大连接时长	60分钟	`aurora dsql connection limits`
单表最大索引数	24	`aurora dsql index limits`
单索引最大列数	8	`aurora dsql index limits`
IDENTITY/SEQUENCE缓存值	1或>=65536	`aurora dsql sequence cache`
支持的列数据类型	查看文档	`aurora dsql supported data types`

验证时机： 在推荐批处理大小、连接池配置或可能触发限制的模式设计前；任何精确数值会影响用户决策的场景。

备选方案： 如果

awsknowledge

不可用，请使用上述默认值，并提示用户应参考DSQL官方文档验证限制。

CLI Tools

CLI工具

Use the AWS CLI,

psql

, and

aurora-dsql-loader

directly for cluster management, SQL connections, and bulk data loading from local or S3 CSV/TSV/Parquet files. Load data-loading.md for loader guidance.

可直接使用AWS CLI、

psql

和

aurora-dsql-loader

进行集群管理、SQL连接，以及从本地或S3加载CSV/TSV/Parquet格式的批量数据。加载data-loading.md获取加载器相关指导。

Quick Start

快速入门

Explore: Use
```
readonly_query
```
with
```
information_schema
```
to list tables. Use
```
get_schema
```
for table structure.
Query: Use
```
readonly_query
```
for SELECT queries. MUST include
```
tenant_id
```
in WHERE for multi-tenant apps. MUST build SQL with
```
safe_query.build()
```
.
Schema changes: Use
```
transact
```
with one DDL per transaction. MUST batch DML under 3,000 rows. MUST use
```
CREATE INDEX ASYNC
```
in a separate call. Use
```
dsql_lint
```
to validate first.
Bulk load data: Use
```
aurora-dsql-loader
```
for CSV/TSV/Parquet. Load data-loading.md for details. Use
```
--dry-run
```
first.

探索： 使用
```
readonly_query
```
查询
```
information_schema
```
列出表。使用
```
get_schema
```
获取表结构。
查询： 使用
```
readonly_query
```
执行SELECT查询。必须在WHERE子句中包含
```
tenant_id
```
以支持多租户应用。必须使用
```
safe_query.build()
```
构建SQL。
模式变更： 使用
```
transact
```
执行单条DDL语句。必须将DML操作分批为不超过3000行。必须在单独调用中使用
```
CREATE INDEX ASYNC
```
。执行前请使用
```
dsql_lint
```
验证。
批量加载数据： 使用
```
aurora-dsql-loader
```
加载CSV/TSV/Parquet数据。详情请查看data-loading.md。请先使用
```
--dry-run
```
进行预验证。

Common Workflows

常见工作流

Workflow 1: Create Multi-Tenant Schema

工作流1：创建多租户模式

Create main table with tenant_id column using transact
Create async index on tenant_id in separate transact call
Create composite indexes for common query patterns (separate transact calls)
Verify schema with get_schema

MUST include tenant_id in all tables
MUST use
```
CREATE INDEX ASYNC
```
exclusively
MUST issue each DDL in its own transact call:
```
transact(["CREATE TABLE ..."])
```
MUST serialize arrays as JSONB; expand at query time with
```
jsonb_array_elements_text(data)
```

使用transact创建包含tenant_id列的主表
在单独的transact调用中创建tenant_id的异步索引
为常见查询模式创建复合索引（单独的transact调用）
使用get_schema验证模式

必须在所有表中包含tenant_id
必须仅使用
```
CREATE INDEX ASYNC
```
必须在单独的transact调用中执行每条DDL：
```
transact(["CREATE TABLE ..."])
```
必须将数组序列化为JSONB；查询时使用
```
jsonb_array_elements_text(data)
```
展开

Workflow 2: Safe Data Migration

工作流2：安全数据迁移

MUST validate every DDL with

dsql_lint(fix=true)

before executing. DML does not require linting.

Validate DDL with
```
dsql_lint(sql=..., fix=true)
```
— handle diagnostics per dsql-lint.md

Add column:

transact(["ALTER TABLE ... ADD COLUMN ..."])

Populate existing rows with UPDATE (batched under 3,000 rows)
Verify with readonly_query COUNT
Create index if needed: validate then
```
transact(["CREATE INDEX ASYNC ..."])
```

MUST issue each
```
ALTER TABLE
```
in its own
```
transact
```
call — DSQL rejects multi-DDL transactions with
```
multiple ddl statements not supported in a transaction
```
MUST add column with only name and type; apply DEFAULT via separate UPDATE
MUST batch updates under 3,000 rows in separate transact calls

Recovery: Resume failed batches by filtering

WHERE new_column IS NULL

执行前必须使用

dsql_lint(fix=true)

验证每条DDL。DML无需验证。

使用
```
dsql_lint(sql=..., fix=true)
```
验证DDL——根据dsql-lint.md处理诊断结果

添加列：

transact(["ALTER TABLE ... ADD COLUMN ..."])

使用UPDATE填充现有行（分批为不超过3000行）
使用readonly_query COUNT验证
如需创建索引：先验证再执行
```
transact(["CREATE INDEX ASYNC ..."])
```

必须在单独的
```
transact
```
调用中执行每条
```
ALTER TABLE
```
——DSQL会拒绝包含多条DDL的事务，提示
```
multiple ddl statements not supported in a transaction
```
必须仅指定列名和类型来添加列；通过单独的UPDATE设置DEFAULT
必须将更新操作分批为不超过3000行，在单独的transact调用中执行

恢复： 通过过滤

WHERE new_column IS NULL

恢复失败的批处理。

Workflow 3: Bulk Data Loading

工作流3：批量数据加载

Use

aurora-dsql-loader

for CSV, TSV, or Parquet loads. MUST load data-loading.md before advising on throughput or diagnosing slow loads.

Validate with
```
--dry-run
```
first
Run with
```
--manifest-dir
```
on persistent storage (not
```
/tmp
```
— tmpfs on AL2023, lost on crash) and
```
--header
```
if file has a header row
On failure: resume with
```
--resume-job-id
```
; for duplicates use
```
--on-conflict do-nothing
```
For large tables: create secondary indexes after load using
```
CREATE INDEX ASYNC
```

使用

aurora-dsql-loader

加载CSV、TSV或Parquet数据。提供吞吐量建议或诊断慢加载问题前必须加载data-loading.md。

先使用
```
--dry-run
```
验证
使用
```
--manifest-dir
```
指定持久化存储（不要使用
```
/tmp
```
——AL2023的tmpfs会在崩溃后丢失数据），如果文件包含表头则添加
```
--header
```
参数
加载失败时：使用
```
--resume-job-id
```
恢复；遇到重复数据时使用
```
--on-conflict do-nothing
```
处理大表时：加载完成后使用
```
CREATE INDEX ASYNC
```
创建二级索引

Workflow 4: Application-Layer Referential Integrity

工作流4：应用层参照完整性

INSERT: MUST validate parent exists with readonly_query → throw error if not found → insert child with transact.

DELETE: MUST check dependents with readonly_query COUNT → return error if dependents exist → delete with transact if safe.

插入： 必须使用readonly_query验证父记录存在→如果未找到则抛出错误→使用transact插入子记录。

删除： 必须使用readonly_query COUNT检查依赖记录→如果存在则返回错误→安全情况下使用transact删除。

Workflow 5: Query with Tenant Isolation

工作流5：租户隔离查询

MUST authorize the caller against the tenant — format validation does not establish authorization
MUST build SQL with
```
safe_query.build()
```
— use
```
allow()
```
/
```
regex()
```
for values (emits
```
'v'
```
),
```
ident()
```
for table/column names (emits
```
"v"
```
). See input-validation.md
MUST include
```
tenant_id
```
in the WHERE clause; reject cross-tenant access at the application layer

必须针对租户对调用者进行授权——格式验证不代表已完成授权
必须使用
```
safe_query.build()
```
构建SQL——对值使用
```
allow()
```
/
```
regex()
```
（生成
```
'v'
```
），对表/列名使用
```
ident()
```
（生成
```
"v"
```
）。详情请查看input-validation.md
必须在WHERE子句中包含
```
tenant_id
```
；在应用层拒绝跨租户访问

Workflow 6: Set Up Scoped Database Roles

工作流6：配置限定范围数据库角色

MUST load access-control.md for role setup, IAM mapping, and schema permissions.

配置角色、IAM映射及模式权限前必须加载access-control.md。

Workflow 7: Table Recreation DDL Migration

工作流7：表重建式DDL迁移

Use the Table Recreation Pattern for

ALTER COLUMN TYPE

DROP COLUMN

DROP CONSTRAINT

, or

MODIFY PRIMARY KEY

. This is a destructive workflow that requires user confirmation at each step. Every generated DDL in the pattern (CREATE new, INSERT ... SELECT, DROP old, RENAME) MUST be validated with

dsql_lint(sql=..., fix=true)

before execution.

MUST load ddl-migrations/overview.md before attempting any of these operations.

针对

ALTER COLUMN TYPE

、

DROP COLUMN

、

DROP CONSTRAINT

或

MODIFY PRIMARY KEY

操作，使用表重建模式。这是一个破坏性工作流，每一步都需要用户确认。模式中生成的每条DDL（创建新表、INSERT ... SELECT、删除旧表、重命名）执行前必须使用

dsql_lint(sql=..., fix=true)

验证。

执行任何此类操作前必须加载ddl-migrations/overview.md。

Workflow 8: Validate and Migrate to DSQL

工作流8：验证并迁移到DSQL

MUST load dsql-lint.md before running

dsql_lint

. Run

dsql_lint(sql=source_sql, fix=true)

to validate and auto-convert. For MySQL-origin SQL, MUST cross-check against mysql-migrations/type-mapping.md even when lint returns clean. On

parse_error

, fall back to manual conversion then re-lint.

运行

dsql_lint

前必须加载dsql-lint.md。运行

dsql_lint(sql=source_sql, fix=true)

进行验证并自动转换。对于源自MySQL的SQL，即使lint结果正常，必须对照mysql-migrations/type-mapping.md进行交叉检查。遇到

parse_error

时，先手动转换再重新lint。

Workflow 9: Query Plan Explainability

工作流9：查询计划可解释性

Explains why the DSQL optimizer chose a particular plan. Triggered by slow queries, high DPU, unexpected Full Scans, or plans the user doesn't understand. REQUIRES a structured Markdown diagnostic report as the deliverable.

MUST load all four reference files at Phase 0: query-plan/plan-interpretation.md, query-plan/catalog-queries.md, query-plan/guc-experiments.md, query-plan/report-format.md. The phase procedures (capture plan, gather evidence, experiment, produce report) are defined in those files.

Safety. Plan capture uses

readonly_query

exclusively. Rewrite DML to SELECT for plan capture. MUST NOT use

transact --allow-writes

for plan capture.

解释DSQL优化器选择特定计划的原因。当遇到慢查询、高DPU、意外全表扫描或用户无法理解的计划时触发。必须交付结构化Markdown诊断报告作为成果。

阶段0时必须加载所有四个参考文档：query-plan/plan-interpretation.md、query-plan/catalog-queries.md、query-plan/guc-experiments.md、query-plan/report-format.md。阶段流程（捕获计划、收集证据、实验、生成报告）在这些文档中定义。

安全注意事项。 计划捕获仅使用

readonly_query

。将DML重写为SELECT进行计划捕获。禁止使用

transact --allow-writes

进行计划捕获。

Workflow 10: Full PostgreSQL → DSQL Schema Migration

工作流10：完整PostgreSQL → DSQL模式迁移

MUST load pg-migrations/type-mapping.md and pg-migrations/schema-objects.md. Run

dsql_lint(fix=true)

first for mechanical fixes, then apply semantic conversions from the pg-migrations references for unfixable diagnostics and patterns the linter cannot handle. Re-lint the final output before deploying.

必须加载pg-migrations/type-mapping.md和pg-migrations/schema-objects.md。先运行

dsql_lint(fix=true)

进行机械修复，然后根据pg-migrations参考文档进行语义转换，处理无法修复的诊断结果和linter无法处理的模式。部署前重新lint最终输出。

Workflow 11: ORM Migration (Django/Hibernate/Rails)

工作流11：ORM迁移（Django/Hibernate/Rails）

Load orm-guides/overview.md for adapter names and framework-specific gotchas.

加载orm-guides/overview.md获取适配器名称及框架特定注意事项。

Error Scenarios

错误场景

awsknowledge
returns no results: Use the default limits in the table above and note that limits should be verified against DSQL documentation.
dsql_lint
unavailable or timing out: See the Error Handling section of dsql-lint.md. Do not silently skip validation — inform the user and require explicit confirmation before proceeding with manual rules from development-guide.md.
OCC serialization error: Retry the transaction. If persistent, check for hot-key contention — see troubleshooting.md.
Transaction exceeds limits: Split into batches under 3,000 rows — see batched-migration.md.
Token expiration mid-operation: Generate a fresh IAM token — see authentication-guide.md. See troubleshooting.md for other issues.

awsknowledge
无返回结果：使用上表中的默认限制，并提示用户应参考DSQL官方文档验证限制。
dsql_lint
不可用或超时：查看dsql-lint.md的错误处理部分。请勿跳过验证——告知用户并要求明确确认后，再根据development-guide.md的手动规则进行操作。
OCC序列化错误： 重试事务。如果持续出现，请检查热键冲突——查看troubleshooting.md。
事务超出限制： 拆分为不超过3000行的批处理——查看batched-migration.md。
操作中令牌过期： 生成新的IAM令牌——查看authentication-guide.md。其他问题请查看troubleshooting.md。