When I alter a table to add columns or change their properties, either in
Enterprise Manager or with T-SQL statements in Query Analyzer, the
alterations are not automatically noticed by the views that use the table.
The only thing I have been able to figure out to do is to find each
potentially affected view and then:
1. Open it in design mode
2. Check or uncheck a field in the Altered table
3. Uncheck or check the same field to restore the view to its original field
selection
4. Close the view and Save it.
Two problems:
1. That's tedious and time-consuming
2. It is easy to miss an affected view.
Can someone advise me as to a better way? (I am mostly using SQL Server 2000
)
Thanks,
Doug MacLeanDoug MacLean wrote:
> When I alter a table to add columns or change their properties, either in
> Enterprise Manager or with T-SQL statements in Query Analyzer, the
> alterations are not automatically noticed by the views that use the table.
> The only thing I have been able to figure out to do is to find each
> potentially affected view and then:
> 1. Open it in design mode
> 2. Check or uncheck a field in the Altered table
> 3. Uncheck or check the same field to restore the view to its original fie
ld
> selection
> 4. Close the view and Save it.
> Two problems:
> 1. That's tedious and time-consuming
> 2. It is easy to miss an affected view.
> Can someone advise me as to a better way? (I am mostly using SQL Server 20
00)
> --
> Thanks,
> Doug MacLean
Firstly, do not use SELECT * in views. It's generally a bad idea to use
SELECT * anywhere in production code. It's a particularly bad idea in
views because of the way views handle changes to the columns.
So assuming you have named columns in all your views, use
sp_refreshview to ensure the view is up to date with changes to its
columns. If you want to add a column you'll have to edit the view
definition of course, which is good practice and not difficult if you
have adequate source control and change control procedures. Table
changes should never be executed by Enterprise Manager in a production
environment.
David Portas, SQL Server MVP
Whenever possible please post enough code to reproduce your problem.
Including CREATE TABLE and INSERT statements usually helps.
State what version of SQL Server you are using and specify the content
of any error messages.
SQL Server Books Online:
http://msdn2.microsoft.com/library/ms130214(en-US,SQL.90).aspx
--|||David,
Thanks. That's very helpful. It eliminates the need for my 4-step process in
Enterprise Manager.
I wonder if there is a tool to list all of the views that depend on a
specified table, because the other challenge is actually finding all of them
when altering a table.
Best Regards,
--
Doug MacLean
"David Portas" wrote:
> Doug MacLean wrote:
> Firstly, do not use SELECT * in views. It's generally a bad idea to use
> SELECT * anywhere in production code. It's a particularly bad idea in
> views because of the way views handle changes to the columns.
> So assuming you have named columns in all your views, use
> sp_refreshview to ensure the view is up to date with changes to its
> columns. If you want to add a column you'll have to edit the view
> definition of course, which is good practice and not difficult if you
> have adequate source control and change control procedures. Table
> changes should never be executed by Enterprise Manager in a production
> environment.
> --
> David Portas, SQL Server MVP
> Whenever possible please post enough code to reproduce your problem.
> Including CREATE TABLE and INSERT statements usually helps.
> State what version of SQL Server you are using and specify the content
> of any error messages.
> SQL Server Books Online:
> http://msdn2.microsoft.com/library/ms130214(en-US,SQL.90).aspx
> --
>|||Greetings,
"Doug MacLean" <DougMacLean@.discussions.microsoft.com> wrote in message
news:C701B020-919E-44B3-A61E-1FCF7AC6AA8D@.microsoft.com...
> David,
> Thanks. That's very helpful. It eliminates the need for my 4-step process
> in
> Enterprise Manager.
> I wonder if there is a tool to list all of the views that depend on a
> specified table, because the other challenge is actually finding all of
> them
> when altering a table.
> Best Regards,
> --
> Doug MacLean
Create the views WITH SCHEMABINDING and you won't be able to change the
table without first dropping the view. This is a good reminder to fix the
view ;-)
Regards,
Neale NOON|||You may want to use this.
select * from INFORMATION_SCHEMA.VIEW_TABLE_USAGE
where table_name = '<table_name>'
-Omnibuzz (The SQL GC)
http://omnibuzz-sql.blogspot.com/|||> I wonder if there is a tool to list all of the views that depend on a
> specified table, because the other challenge is actually finding all of
> them
> when altering a table.
You might find it easier to refresh all views since you cannot rely on
dependency information unless the views were created WITH SCHEMABINDING.
The script below will generate a script to refresh all views in the current
database. You can wrap it in a cursor and execute according to your
preference. BTW, even without 'SELECT *', you can run into issues with
changed datatypes in the referenced tables.
SELECT
'EXEC sp_refreshview ''' +
QUOTENAME(TABLE_SCHEMA) +
N'.' +
QUOTENAME(TABLE_NAME) +
''''
FROM INFORMATION_SCHEMA.TABLES
WHERE TABLE_TYPE = 'VIEW'
AND OBJECTPROPERTY(OBJECT_ID(
QUOTENAME(TABLE_SCHEMA) +
N'.' +
QUOTENAME(TABLE_NAME)
), 'IsMSShipped') = 0
Hope this helps.
Dan Guzman
SQL Server MVP
"Doug MacLean" <DougMacLean@.discussions.microsoft.com> wrote in message
news:C701B020-919E-44B3-A61E-1FCF7AC6AA8D@.microsoft.com...
> David,
> Thanks. That's very helpful. It eliminates the need for my 4-step process
> in
> Enterprise Manager.
> I wonder if there is a tool to list all of the views that depend on a
> specified table, because the other challenge is actually finding all of
> them
> when altering a table.
> Best Regards,
> --
> Doug MacLean
>
> "David Portas" wrote:
>|||Thanks, Neale.
That's an interesting idea to know when I change a table that has one or
more dependent views. So it is a good reminder.
But I KNOW there are views. And what I want is a clean way of finding and
refreshing them. Adding a process of dropping (and then re-adding) them seem
s
tedious.
Thanks Best Regards,
--
Doug MacLean
"Neale NOON" wrote:
> Greetings,
> "Doug MacLean" <DougMacLean@.discussions.microsoft.com> wrote in message
> news:C701B020-919E-44B3-A61E-1FCF7AC6AA8D@.microsoft.com...
>
> Create the views WITH SCHEMABINDING and you won't be able to change the
> table without first dropping the view. This is a good reminder to fix the
> view ;-)
> --
> Regards,
> Neale NOON
>
>|||Dan,
Ahhh!. That's a great idea. Thanks for both the idea and the sample code.
And it's practical because I only do such updates on the production system a
t
times when no one is using it.
You're right. I've already seen the problem that -- even without adding
columns -- the view needs to be refreshed.
Thanks much and Best regards,
--
Doug MacLean
"Dan Guzman" wrote:
> You might find it easier to refresh all views since you cannot rely on
> dependency information unless the views were created WITH SCHEMABINDING.
> The script below will generate a script to refresh all views in the curren
t
> database. You can wrap it in a cursor and execute according to your
> preference. BTW, even without 'SELECT *', you can run into issues with
> changed datatypes in the referenced tables.
> SELECT
> 'EXEC sp_refreshview ''' +
> QUOTENAME(TABLE_SCHEMA) +
> N'.' +
> QUOTENAME(TABLE_NAME) +
> ''''
> FROM INFORMATION_SCHEMA.TABLES
> WHERE TABLE_TYPE = 'VIEW'
> AND OBJECTPROPERTY(OBJECT_ID(
> QUOTENAME(TABLE_SCHEMA) +
> N'.' +
> QUOTENAME(TABLE_NAME)
> ), 'IsMSShipped') = 0
> --
> Hope this helps.
> Dan Guzman
> SQL Server MVP
> "Doug MacLean" <DougMacLean@.discussions.microsoft.com> wrote in message
> news:C701B020-919E-44B3-A61E-1FCF7AC6AA8D@.microsoft.com...
>
>
