SQL Server Migration Assistant for MySQL(SSMA),数据迁移后的一些善后操作

微软的SQL Server Migration Assistant for MySQL(此后简称SSMA)用于MySQL迁移非常好用,如何使用,网上大把,本篇原创博客旨在善后工作。


迁移之后,需要把ssma$rowid这字段删除,该字段的作用就是给每个行做一个GUID的唯一标识,用以迁移数据使用。


注意:以下脚本均以既定的格式为例,自定义的约束名、索引名等无法识别。


step 1: 由于ssma$rowid拥有唯一约束,在删除这一列前,必须将约束删除

--删除rowid的唯一约束
select 'ALTER TABLE ['+t.name+'] DROP CONSTRAINT '+i.name from sys.indexes i
inner join sys.tables t on i.object_id=t.object_id
where i.type_desc='NONCLUSTERED'
and i.name like '%ssma$rowid'


step 2: 唯一约束删除后,将ssma$rowid列删除

--删除rowid这一列
select distinct 'ALTER TABLE ['+t.name+'] DROP column ssma$rowid ' from sys.tables t
left join sys.columns c on t.object_id=t.object_id
where t.name <> '__MigrationHistory' and c.name = 'ssma$rowid'



如果是GUID式的主键,迁移后的数据类型为nchar(36),需要更改类型为uniqueidentifier。

在更改前,需要将主键约束去除。

而主键约束会被其他表的外键引用,得先把外键约束去除。

step 1: 

--删除每个表的外键约束
select 'ALTER TABLE ['+t.name+'] DROP CONSTRAINT '+o.name from sys.objects o
inner join sys.tables t on o.parent_object_id=t.object_id
where o.type_desc='FOREIGN_KEY_CONSTRAINT'
and o.name like '%$FK_%'

step 2:

--删除每个表的主键约束
select 'ALTER TABLE ['+t.name+'] DROP CONSTRAINT '+i.name from sys.indexes i
inner join sys.tables t on i.object_id=t.object_id
where i.type_desc='CLUSTERED'
and i.name like '%_Id'


step 3:

--删除每个表的默认值约束
select 'ALTER TABLE ['+t.name+'] DROP CONSTRAINT '+o.name from sys.objects o
inner join sys.tables t on o.parent_object_id=t.object_id
where o.type_desc='DEFAULT_CONSTRAINT'
and o.name like 'DF_%'

step 4:

--删除每个表的索引
select 'DROP INDEX ['+i.name+'] ON ['+t.name+'] WITH ( ONLINE = OFF )' from sys.indexes i
inner join sys.tables t on i.object_id=t.object_id
where i.type_desc='NONCLUSTERED'
and i.name like 'IX_%'

step 5:

/*GUID & pk*/
select 
(case Is_Nullable  when 'NO'
 then concat('alter table [',table_name,'] alter column ',column_name,' uniqueidentifier not null')
else concat('alter table [',table_name,'] alter column ',column_name,' uniqueidentifier') end) 
 from INFORMATION_SCHEMA.COLUMNS where COLUMN_NAME like '%Id' and DATA_TYPE='nchar' and CHARACTER_MAXIMUM_LENGTH=36;



一般稍有规模的项目都会有并发控制,做法一般为RowVersion

迁移后的数据类型为datetime,需要更改为timestamp

/*RowVersion*/
-- 添加新列 RowVersion1
select concat('alter table [',table_name,'] add ',column_name,'1 timestamp not null') from INFORMATION_SCHEMA.COLUMNS where DATA_TYPE='datetime' and column_name='rowVersion'

-- 删除旧的RowVersion列
select concat('alter table [',table_name,'] drop column [',column_name,']') from INFORMATION_SCHEMA.COLUMNS where DATA_TYPE='datetime' and column_name='rowVersion'

--重命名RowVersion1=>RowVersion
select 'exec sp_rename '''+TABLE_NAME+'.'+COLUMN_NAME+''','''+'RowVersion'+''',''column''' from INFORMATION_SCHEMA.COLUMNS where DATA_TYPE='timestamp' and column_name='rowVersion1'


后记:这些约束在更改数据类型的时候,会碍事,在更改完成后,自然需重新加上。

谢谢观看



你可能感兴趣的:(数据库)