Thread: Prob. with migration plugin wrt. MS SQL source DB
Hello, in trying to run the migration tool on a fairly large MS SQL DB I am receiving the following error:
-------------------------------------
Creating table: NDF__COMPANY_MARKETING_CONTACTS
Creating sequence NDF__COMPANY_MARKETING_CONTACTS_NDF__ID_key
Creating sequence NDF__COMPANY_MARKETING_CONTACTS_NDF__ID_key
An error occured at: 20/11/2002 12:27:47 PM:
-2147467259: ERROR: Relation 'NDF__COMPANY_MARKETING_CONTACTS' already exists
-2147467259: ERROR: Relation 'NDF__COMPANY_MARKETING_CONTACTS' already exists
Rolling back... Done.
-------------------------------------
From MS SQL (2000) the SQL of the table is as follows:
-------------------------------------
CREATE TABLE [dbo].[NDF__COMPANY_MARKETING_CONTACTS] (
[NDF__ID] [int] IDENTITY (1, 1) NOT NULL ,
[CONTACTS_ID] [int] NULL ,
[COMPANIES_ID] [int] NULL ,
[NDF_VALUE] [varchar] (100) COLLATE SQL_Latin1_General_CP1_CI_AS NULL
) ON [PRIMARY]
GO
[NDF__ID] [int] IDENTITY (1, 1) NOT NULL ,
[CONTACTS_ID] [int] NULL ,
[COMPANIES_ID] [int] NULL ,
[NDF_VALUE] [varchar] (100) COLLATE SQL_Latin1_General_CP1_CI_AS NULL
) ON [PRIMARY]
GO
ALTER TABLE [dbo].[NDF__COMPANY_MARKETING_CONTACTS] WITH NOCHECK ADD
CONSTRAINT [PK_NDF__COMPANY_MARKETING_CONTACTS] PRIMARY KEY CLUSTERED
(
[NDF__ID]
) ON [PRIMARY]
GO
CONSTRAINT [PK_NDF__COMPANY_MARKETING_CONTACTS] PRIMARY KEY CLUSTERED
(
[NDF__ID]
) ON [PRIMARY]
GO
-------------------------------------
It seems like there might be some limitation on the length of a relation's identifier. That is, it looks as if the sequence name is getting truncated.
Any thoughts? Is this something with the migration plugin or pgadmin or postgresql??
Thanks for any help!
Jason.
Hi Jason.
In PostgreSQL 7.2.x and below, the name datatype (used by identifiers) is 32 characters long. This means that names cannot exceed 31 characters (1 char is used for a terminating null). In PostgreSQL 7.3 (to be released any day now), this limit has been increased to 64 (practical length 63). You can hack 7.2.x by editting the NAMEDATALEN value which is in src/include/postgres.h (iirc) and recompiling. This does require an initdb though, and may affect performance. For further info on this checkout the postgresql list archives (archives.postgresql.org).
Regards, Dave.
-----Original Message-----
From: Jason Uberig [mailto:uberjay@hotmail.com]
Sent: 20 November 2002 18:29
To: pgadmin-support@postgresql.org
Subject: [pgadmin-support] Prob. with migration plugin wrt. MS SQL source DBHello, in trying to run the migration tool on a fairly large MS SQL DB I am receiving the following error:-------------------------------------Creating table: NDF__COMPANY_MARKETING_CONTACTS
Creating sequence NDF__COMPANY_MARKETING_CONTACTS_NDF__ID_keyAn error occured at: 20/11/2002 12:27:47 PM:
-2147467259: ERROR: Relation 'NDF__COMPANY_MARKETING_CONTACTS' already existsRolling back... Done.-------------------------------------From MS SQL (2000) the SQL of the table is as follows:-------------------------------------CREATE TABLE [dbo].[NDF__COMPANY_MARKETING_CONTACTS] (
[NDF__ID] [int] IDENTITY (1, 1) NOT NULL ,
[CONTACTS_ID] [int] NULL ,
[COMPANIES_ID] [int] NULL ,
[NDF_VALUE] [varchar] (100) COLLATE SQL_Latin1_General_CP1_CI_AS NULL
) ON [PRIMARY]
GOALTER TABLE [dbo].[NDF__COMPANY_MARKETING_CONTACTS] WITH NOCHECK ADD
CONSTRAINT [PK_NDF__COMPANY_MARKETING_CONTACTS] PRIMARY KEY CLUSTERED
(
[NDF__ID]
) ON [PRIMARY]
GO-------------------------------------It seems like there might be some limitation on the length of a relation's identifier. That is, it looks as if the sequence name is getting truncated.Any thoughts? Is this something with the migration plugin or pgadmin or postgresql??Thanks for any help!Jason.