database collation: multiple

Showing posts with label multiple. Show all posts

Thursday, March 29, 2012

Changing the text of the code of multiple stored procedure

Hi,
We have four different SQL Servers and each server has 2-3 databases.
The stored procedures on each database use link servers to access each
other. The link server names are based on the actual server name.
The Issue:- We are migrating to SQL 2005 and to save on the licence
cost we are planning to consolidate all our databases on one more
powerful machine which can take load of peak hours. Now will it be
advisable to create link servers on the new server by same names as
existing on the current servers, so that we don't have to change the
code. Or should we change the code and remove link servers from in the
code of each SP and trigger.
I think quires using the link server will be slower then directly
assessing other databases.
For Ex <linkserver>.<dbname>.dbo.<tablename> vs. <dbname>..<tablename>
If the best solution is to change the code then what is the best &
fast way to do so (there are more then 1000 SP, functions and triggers
to be changed).
Following are the few differrent approches I've followed:-
1) I tried to change in the syscomments table but SQL Server 2005
won't allow me to change the sys tables.
2) I wrote a SP to find out the search string in the code using
systables and once I had the object name where the search string is
used then i imported the code in a temp table using sp_helptext. Then
removed the linkserver in the temp table and bcp the temp table into a
txt file (upto this point I have created a script of SP/trigger with
the changed code)
The issue:- a) BCP is inserting multiple blank lines in the txt file.
How do I make sure these unwanted blank lines do not appear in the txt
file. The reason why I'm more worried abt the blank lines is because
each SP might go thru multiple change cycles as I described above (if
different linkservers are used in the same SP/trigger) and every time
there will be balnak lines added which will blot the code in terms of
LOC and makes it difficult to maintain long term (readability issues).
The bcp command I'm using as follows:-
'bcp '+ @.db_name + '.dbo.temp_helptext out '+ @.bcp_path + '' +
@.obj_name + '.txt -c -U<xx> -P<xxxxx> -S<xxxxxxxxxx>'
b) How do I execute 1000 or so scripts automatically.
Thanks very much in advance to respond.
Cheers,
SGHi
"subodh97@.gmail.com" wrote:
> Hi,
> We have four different SQL Servers and each server has 2-3 databases.
> The stored procedures on each database use link servers to access each
> other. The link server names are based on the actual server name.
> The Issue:- We are migrating to SQL 2005 and to save on the licence
> cost we are planning to consolidate all our databases on one more
> powerful machine which can take load of peak hours. Now will it be
> advisable to create link servers on the new server by same names as
> existing on the current servers, so that we don't have to change the
> code. Or should we change the code and remove link servers from in the
> code of each SP and trigger.
> I think quires using the link server will be slower then directly
> assessing other databases.
> For Ex <linkserver>.<dbname>.dbo.<tablename> vs. <dbname>..<tablename>
> If the best solution is to change the code then what is the best &
> fast way to do so (there are more then 1000 SP, functions and triggers
> to be changed).
> Following are the few differrent approches I've followed:-
> 1) I tried to change in the syscomments table but SQL Server 2005
> won't allow me to change the sys tables.
> 2) I wrote a SP to find out the search string in the code using
> systables and once I had the object name where the search string is
> used then i imported the code in a temp table using sp_helptext. Then
> removed the linkserver in the temp table and bcp the temp table into a
> txt file (upto this point I have created a script of SP/trigger with
> the changed code)
> The issue:- a) BCP is inserting multiple blank lines in the txt file.
> How do I make sure these unwanted blank lines do not appear in the txt
> file. The reason why I'm more worried abt the blank lines is because
> each SP might go thru multiple change cycles as I described above (if
> different linkservers are used in the same SP/trigger) and every time
> there will be balnak lines added which will blot the code in terms of
> LOC and makes it difficult to maintain long term (readability issues).
> The bcp command I'm using as follows:-
> 'bcp '+ @.db_name + '.dbo.temp_helptext out '+ @.bcp_path + '' +
> @.obj_name + '.txt -c -U<xx> -P<xxxxx> -S<xxxxxxxxxx>'
> b) How do I execute 1000 or so scripts automatically.
> Thanks very much in advance to respond.
> Cheers,
> SG
>
The easiest way would be to script out your stored procedures using the
scripting options in SSMS or Enterprise Manager and then globally replace
each instance with a text editor. If you used source code control for the
your database code this would not be an issue and you can track and control
the changes to the source when you start to implement SQL 2005 features.
John|||SG,
I'm not sure about SQL Server 2005, but with SQL Server 2K I remember having
performance overhead when using a 4 part name with a database on the same
server. SQL Server still goes through unnecessary additional security,
collation conversion, etc. layers as if it was linking to another server.
You might want to consider the following low tech approach:
1. Script out all stored procs to a file.
2. Replace "ServerName." with "".
3. Replace "Create Procedure" with "Alter Procedure".
4. Apply the script.
-- Bill
<subodh97@.gmail.com> wrote in message
news:1173977536.347970.223200@.p15g2000hsd.googlegroups.com...
> Hi,
> We have four different SQL Servers and each server has 2-3 databases.
> The stored procedures on each database use link servers to access each
> other. The link server names are based on the actual server name.
> The Issue:- We are migrating to SQL 2005 and to save on the licence
> cost we are planning to consolidate all our databases on one more
> powerful machine which can take load of peak hours. Now will it be
> advisable to create link servers on the new server by same names as
> existing on the current servers, so that we don't have to change the
> code. Or should we change the code and remove link servers from in the
> code of each SP and trigger.
> I think quires using the link server will be slower then directly
> assessing other databases.
> For Ex <linkserver>.<dbname>.dbo.<tablename> vs. <dbname>..<tablename>
> If the best solution is to change the code then what is the best &
> fast way to do so (there are more then 1000 SP, functions and triggers
> to be changed).
> Following are the few differrent approches I've followed:-
> 1) I tried to change in the syscomments table but SQL Server 2005
> won't allow me to change the sys tables.
> 2) I wrote a SP to find out the search string in the code using
> systables and once I had the object name where the search string is
> used then i imported the code in a temp table using sp_helptext. Then
> removed the linkserver in the temp table and bcp the temp table into a
> txt file (upto this point I have created a script of SP/trigger with
> the changed code)
> The issue:- a) BCP is inserting multiple blank lines in the txt file.
> How do I make sure these unwanted blank lines do not appear in the txt
> file. The reason why I'm more worried abt the blank lines is because
> each SP might go thru multiple change cycles as I described above (if
> different linkservers are used in the same SP/trigger) and every time
> there will be balnak lines added which will blot the code in terms of
> LOC and makes it difficult to maintain long term (readability issues).
> The bcp command I'm using as follows:-
> 'bcp '+ @.db_name + '.dbo.temp_helptext out '+ @.bcp_path + '' +
> @.obj_name + '.txt -c -U<xx> -P<xxxxx> -S<xxxxxxxxxx>'
> b) How do I execute 1000 or so scripts automatically.
> Thanks very much in advance to respond.
> Cheers,
> SG
>|||On Mar 15, 1:39 pm, John Bell <jbellnewspo...@.hotmail.com> wrote:
> Hi
>
> "subod...@.gmail.com" wrote:
> > Hi,
> > We have four different SQL Servers and each server has 2-3 databases.
> > The stored procedures on each database use link servers to access each
> > other. The link server names are based on the actual server name.
> > The Issue:- We are migrating to SQL 2005 and to save on the licence
> > cost we are planning to consolidate all our databases on one more
> > powerful machine which can take load of peak hours. Now will it be
> > advisable to create link servers on the new server by same names as
> > existing on the current servers, so that we don't have to change the
> > code. Or should we change the code and remove link servers from in the
> > code of each SP and trigger.
> > I think quires using the link server will be slower then directly
> > assessing other databases.
> > For Ex <linkserver>.<dbname>.dbo.<tablename> vs. <dbname>..<tablename>
> > If the best solution is to change the code then what is the best &
> > fast way to do so (there are more then 1000 SP, functions and triggers
> > to be changed).
> > Following are the few differrent approches I've followed:-
> > 1) I tried to change in the syscomments table but SQL Server 2005
> > won't allow me to change the sys tables.
> > 2) I wrote a SP to find out the search string in the code using
> > systables and once I had the object name where the search string is
> > used then i imported the code in a temp table using sp_helptext. Then
> > removed the linkserver in the temp table and bcp the temp table into a
> > txt file (upto this point I have created a script of SP/trigger with
> > the changed code)
> > The issue:- a) BCP is inserting multiple blank lines in the txt file.
> > How do I make sure these unwanted blank lines do not appear in the txt
> > file. The reason why I'm more worried abt the blank lines is because
> > each SP might go thru multiple change cycles as I described above (if
> > different linkservers are used in the same SP/trigger) and every time
> > there will be balnak lines added which will blot the code in terms of
> > LOC and makes it difficult to maintain long term (readability issues).
> > The bcp command I'm using as follows:-
> > 'bcp '+ @.db_name + '.dbo.temp_helptext out '+ @.bcp_path + '' +
> > @.obj_name + '.txt -c -U<xx> -P<xxxxx> -S<xxxxxxxxxx>'
> > b) How do I execute 1000 or so scripts automatically.
> > Thanks very much in advance to respond.
> > Cheers,
> > SG
> The easiest way would be to script out your stored procedures using the
> scripting options in SSMS or Enterprise Manager and then globally replace
> each instance with a text editor. If you used source code control for the
> your database code this would not be an issue and you can track and control
> the changes to the source when you start to implement SQL 2005 features.
> John
Thanks John, The thing is we don't have any source control at this
point (may be future we'll have VSS). And also to run the scripts
manually which I don't want to do on the final day of migration. If
somebody can help me telling the way how to get away with the numerous
blank lines while bcp..ing the code to a text file then half of the
battle is won. Following is the actual code of the SP:-
/*
Author - SG
*/
ALTER PROC [dbo].[sp_use_linkserver]
AS
BEGIN
DECLARE
@.count INT
-- Just a comment
SET @.count = (SELECT count(*) FROM LINKSRV2.testdb.dbo.testtable)
print 'Total # of sites in R3 = ' + CAST(@.count AS VARCHAR(10))
END
Following is the code in txt file generated:-
/*
Author - Subodh Goyal
*/
CREATE PROC [dbo].[sp_use_linkserver]
AS
BEGIN
DECLARE
@.count INT
-- Just a comment
SET @.count = (SELECT count(*) FROM LINKSRV1.testdb.dbo.testtable)
print 'Total # of sites in R3 = ' + CAST(@.count AS VARCHAR(10))
END
Hope this help in understanding the issue.
SG|||On Mar 15, 2:33 pm, "AlterEgo" <altereg...@.dslextreme.com> wrote:
> SG,
> I'm not sure about SQL Server 2005, but with SQL Server 2K I remember having
> performance overhead when using a 4 part name with a database on the same
> server. SQL Server still goes through unnecessary additional security,
> collation conversion, etc. layers as if it was linking to another server.
> You might want to consider the following low tech approach:
> 1. Script out all stored procs to a file.
> 2. Replace "ServerName." with "".
> 3. Replace "Create Procedure" with "Alter Procedure".
> 4. Apply the script.
> -- Bill
> <subod...@.gmail.com> wrote in message
> news:1173977536.347970.223200@.p15g2000hsd.googlegroups.com...
> > Hi,
> > We have four different SQL Servers and each server has 2-3 databases.
> > The stored procedures on each database use link servers to access each
> > other. The link server names are based on the actual server name.
> > The Issue:- We are migrating to SQL 2005 and to save on the licence
> > cost we are planning to consolidate all our databases on one more
> > powerful machine which can take load of peak hours. Now will it be
> > advisable to create link servers on the new server by same names as
> > existing on the current servers, so that we don't have to change the
> > code. Or should we change the code and remove link servers from in the
> > code of each SP and trigger.
> > I think quires using the link server will be slower then directly
> > assessing other databases.
> > For Ex <linkserver>.<dbname>.dbo.<tablename> vs. <dbname>..<tablename>
> > If the best solution is to change the code then what is the best &
> > fast way to do so (there are more then 1000 SP, functions and triggers
> > to be changed).
> > Following are the few differrent approches I've followed:-
> > 1) I tried to change in the syscomments table but SQL Server 2005
> > won't allow me to change the sys tables.
> > 2) I wrote a SP to find out the search string in the code using
> > systables and once I had the object name where the search string is
> > used then i imported the code in a temp table using sp_helptext. Then
> > removed the linkserver in the temp table and bcp the temp table into a
> > txt file (upto this point I have created a script of SP/trigger with
> > the changed code)
> > The issue:- a) BCP is inserting multiple blank lines in the txt file.
> > How do I make sure these unwanted blank lines do not appear in the txt
> > file. The reason why I'm more worried abt the blank lines is because
> > each SP might go thru multiple change cycles as I described above (if
> > different linkservers are used in the same SP/trigger) and every time
> > there will be balnak lines added which will blot the code in terms of
> > LOC and makes it difficult to maintain long term (readability issues).
> > The bcp command I'm using as follows:-
> > 'bcp '+ @.db_name + '.dbo.temp_helptext out '+ @.bcp_path + '' +
> > @.obj_name + '.txt -c -U<xx> -P<xxxxx> -S<xxxxxxxxxx>'
> > b) How do I execute 1000 or so scripts automatically.
> > Thanks very much in advance to respond.
> > Cheers,
> > SG
Thanks Bill,
The only issue in following this approach is sheer amount of work on
the final migration day. There are about 12 different databases which
will be migrating on single machine. Also I can not script all the
databases before hand they need to be scripted on the same day. Also
there are SP. functions and triggers which will make this exercise not
smooth. Anyways this is going to be my last option if nothing else can
be done before the final migration day.
SG|||Hi
sp_helptext puts in the extra lines, you could replace the extra char(10)
and/or char(13) characters using the SQL REPLACE function, but I am not sure
that the method you are using is going to work as it may not reliably produce
line breaks in the correct place and you may get words broken.
If you do not want to use scripting then look at using DMO to get the
procedure definition and change them accordingly. This will allow you to only
change the ones where you find the linked servers.
John
"subodh97@.gmail.com" wrote:
> On Mar 15, 1:39 pm, John Bell <jbellnewspo...@.hotmail.com> wrote:
> > Hi
> >
> >
> >
> > "subod...@.gmail.com" wrote:
> > > Hi,
> >
> > > We have four different SQL Servers and each server has 2-3 databases.
> > > The stored procedures on each database use link servers to access each
> > > other. The link server names are based on the actual server name.
> >
> > > The Issue:- We are migrating to SQL 2005 and to save on the licence
> > > cost we are planning to consolidate all our databases on one more
> > > powerful machine which can take load of peak hours. Now will it be
> > > advisable to create link servers on the new server by same names as
> > > existing on the current servers, so that we don't have to change the
> > > code. Or should we change the code and remove link servers from in the
> > > code of each SP and trigger.
> >
> > > I think quires using the link server will be slower then directly
> > > assessing other databases.
> >
> > > For Ex <linkserver>.<dbname>.dbo.<tablename> vs. <dbname>..<tablename>
> >
> > > If the best solution is to change the code then what is the best &
> > > fast way to do so (there are more then 1000 SP, functions and triggers
> > > to be changed).
> >
> > > Following are the few differrent approches I've followed:-
> >
> > > 1) I tried to change in the syscomments table but SQL Server 2005
> > > won't allow me to change the sys tables.
> > > 2) I wrote a SP to find out the search string in the code using
> > > systables and once I had the object name where the search string is
> > > used then i imported the code in a temp table using sp_helptext. Then
> > > removed the linkserver in the temp table and bcp the temp table into a
> > > txt file (upto this point I have created a script of SP/trigger with
> > > the changed code)
> >
> > > The issue:- a) BCP is inserting multiple blank lines in the txt file.
> > > How do I make sure these unwanted blank lines do not appear in the txt
> > > file. The reason why I'm more worried abt the blank lines is because
> > > each SP might go thru multiple change cycles as I described above (if
> > > different linkservers are used in the same SP/trigger) and every time
> > > there will be balnak lines added which will blot the code in terms of
> > > LOC and makes it difficult to maintain long term (readability issues).
> >
> > > The bcp command I'm using as follows:-
> > > 'bcp '+ @.db_name + '.dbo.temp_helptext out '+ @.bcp_path + '' +
> > > @.obj_name + '.txt -c -U<xx> -P<xxxxx> -S<xxxxxxxxxx>'
> >
> > > b) How do I execute 1000 or so scripts automatically.
> >
> > > Thanks very much in advance to respond.
> >
> > > Cheers,
> >
> > > SG
> >
> > The easiest way would be to script out your stored procedures using the
> > scripting options in SSMS or Enterprise Manager and then globally replace
> > each instance with a text editor. If you used source code control for the
> > your database code this would not be an issue and you can track and control
> > the changes to the source when you start to implement SQL 2005 features.
> >
> > John
> Thanks John, The thing is we don't have any source control at this
> point (may be future we'll have VSS). And also to run the scripts
> manually which I don't want to do on the final day of migration. If
> somebody can help me telling the way how to get away with the numerous
> blank lines while bcp..ing the code to a text file then half of the
> battle is won. Following is the actual code of the SP:-
> /*
> Author - SG
> */
> ALTER PROC [dbo].[sp_use_linkserver]
> AS
> BEGIN
> DECLARE
> @.count INT
> -- Just a comment
> SET @.count = (SELECT count(*) FROM LINKSRV2.testdb.dbo.testtable)
> print 'Total # of sites in R3 = ' + CAST(@.count AS VARCHAR(10))
> END
>
> Following is the code in txt file generated:-
>
> /*
> Author - Subodh Goyal
> */
>
> CREATE PROC [dbo].[sp_use_linkserver]
>
> AS
>
> BEGIN
>
> DECLARE
> @.count INT
> -- Just a comment
> SET @.count = (SELECT count(*) FROM LINKSRV1.testdb.dbo.testtable)
> print 'Total # of sites in R3 = ' + CAST(@.count AS VARCHAR(10))
> END
>
>
>
>
>
>
>
> Hope this help in understanding the issue.
> SG
>|||Hi
I am surprised that your stored procedures/functions/triggers are so
volatile you can not do this till the last minute? It is not a good idea to
combine an upgrade with major changes to the software unless it is necessary
to make it work on the new system. It is not a good idea to implement these
changes blindly without doing an test migration and regression test cycle
before you implement this on a live system.
Have you run the SQL 2005 upgrade advisor on these systems?
John|||On Mar 15, 9:52 pm, subod...@.gmail.com wrote:
> Hi,
> We have four different SQL Servers and each server has 2-3 databases.
> The stored procedures on each database use link servers to access each
> other. The link server names are based on the actual server name.
> The Issue:- We are migrating to SQL 2005 and to save on the licence
> cost we are planning to consolidate all our databases on one more
> powerful machine which can take load of peak hours. Now will it be
> advisable to create link servers on the new server by same names as
> existing on the current servers, so that we don't have to change the
> code. Or should we change the code and remove link servers from in the
> code of each SP and trigger.
> I think quires using the link server will be slower then directly
> assessing other databases.
> For Ex <linkserver>.<dbname>.dbo.<tablename> vs. <dbname>..<tablename>
> If the best solution is to change the code then what is the best &
> fast way to do so (there are more then 1000 SP, functions and triggers
> to be changed).
> Following are the few differrent approches I've followed:-
> 1) I tried to change in the syscomments table but SQL Server 2005
> won't allow me to change the sys tables.
> 2) I wrote a SP to find out the search string in the code using
> systables and once I had the object name where the search string is
> used then i imported the code in a temp table using sp_helptext. Then
> removed the linkserver in the temp table and bcp the temp table into a
> txt file (upto this point I have created a script of SP/trigger with
> the changed code)
> The issue:- a) BCP is inserting multiple blank lines in the txt file.
> How do I make sure these unwanted blank lines do not appear in the txt
> file. The reason why I'm more worried abt the blank lines is because
> each SP might go thru multiple change cycles as I described above (if
> different linkservers are used in the same SP/trigger) and every time
> there will be balnak lines added which will blot the code in terms of
> LOC and makes it difficult to maintain long term (readability issues).
> The bcp command I'm using as follows:-
> 'bcp '+ @.db_name + '.dbo.temp_helptext out '+ @.bcp_path + '' +
> @.obj_name + '.txt -c -U<xx> -P<xxxxx> -S<xxxxxxxxxx>'
> b) How do I execute 1000 or so scripts automatically.
> Thanks very much in advance to respond.
> Cheers,
> SG
Please check synonym in BOL|||On Mar 16, 4:56 am, John Bell <jbellnewspo...@.hotmail.com> wrote:
> Hi
> I am surprised that your stored procedures/functions/triggers are so
> volatile you can not do this till the last minute? It is not a good idea to
> combine an upgrade with major changes to the software unless it is necessary
> to make it work on the new system. It is not a good idea to implement these
> changes blindly without doing an test migration and regression test cycle
> before you implement this on a live system.
> Have you run the SQL 2005 upgrade advisor on these systems?
> John
Hi John, Thanks for the reply. We are getting the new hardware for the
migration and planning to do a mock migration first on the new server
and put that new server and applications on testing mode. Our current
production servers will still run in parallel. There will be a
difference of abt a month or so in mock and actual migration and I
cannot stop any release of new code to the current production servers
during that one month or so period due to the nature of business and
4-5 different teams involved in it. The solution I'm looking is to
generate the right scripts dynamically so that I can incorporate any
changes done during the month of testing in the final migration.
By the way I've run Upgrade adviser on all the servers and there were
no major issues other then few of the DTS packages we have to re-write
in 2005 SSIS.
I'll try DMO option to see how effective and easy it can be.
Thanks,
SG|||On Mar 16, 2:07 pm, subod...@.gmail.com wrote:
> On Mar 16, 4:56 am, John Bell <jbellnewspo...@.hotmail.com> wrote:
> Hi John, Thanks for the reply. We are getting the new hardware for the
> migration and planning to do a mock migration first on the new server
> and put that new server and applications on testing mode. Our current
> production servers will still run in parallel. There will be a
> difference of abt a month or so in mock and actual migration and I
> cannot stop any release of new code to the current production servers
> during that one month or so period due to the nature of business and
> 4-5 different teams involved in it. The solution I'm looking is to
> generate the right scripts dynamically so that I can incorporate any
> changes done during the month of testing in the final migration.
>
Not freezing development is making this task significantly more risky.
Source code control would have made your task significantly easier. An
option may be to have a second cut over for a shorter period where you
run through all the tests and freeze the code for that shorter time? A
different approach would be to take backups of the databases that were
cut over and then compare a restored backup against the live database
at the final cutover using a tool such as Red Gate SQL Compare or
DBGhost. This will show you everything that you will need to change in
the migration. Having two copies of a given database may require
significant disc space, but they do not have to be on the same
machines.
> By the way I've run Upgrade adviser on all the servers and there were
> no major issues other then few of the DTS packages we have to re-write
> in 2005 SSIS.
> I'll try DMO option to see how effective and easy it can be.
> Thanks,
> SG
Johnsql

Changing the text of the code of multiple stored procedure

Hi,
We have four different SQL Servers and each server has 2-3 databases.
The stored procedures on each database use link servers to access each
other. The link server names are based on the actual server name.
The Issue:- We are migrating to SQL 2005 and to save on the licence
cost we are planning to consolidate all our databases on one more
powerful machine which can take load of peak hours. Now will it be
advisable to create link servers on the new server by same names as
existing on the current servers, so that we don't have to change the
code. Or should we change the code and remove link servers from in the
code of each SP and trigger.
I think quires using the link server will be slower then directly
assessing other databases.
For Ex <linkserver>.<dbname>.dbo.<tablename> vs. <dbname>..<tablename>
If the best solution is to change the code then what is the best &
fast way to do so (there are more then 1000 SP, functions and triggers
to be changed).
Following are the few differrent approches I've followed:-
1) I tried to change in the syscomments table but SQL Server 2005
won't allow me to change the sys tables.
2) I wrote a SP to find out the search string in the code using
systables and once I had the object name where the search string is
used then i imported the code in a temp table using sp_helptext. Then
removed the linkserver in the temp table and bcp the temp table into a
txt file (upto this point I have created a script of SP/trigger with
the changed code)
The issue:- a) BCP is inserting multiple blank lines in the txt file.
How do I make sure these unwanted blank lines do not appear in the txt
file. The reason why I'm more worried abt the blank lines is because
each SP might go thru multiple change cycles as I described above (if
different linkservers are used in the same SP/trigger) and every time
there will be balnak lines added which will blot the code in terms of
LOC and makes it difficult to maintain long term (readability issues).
The bcp command I'm using as follows:-
'bcp '+ @.db_name + '.dbo.temp_helptext out '+ @.bcp_path + '' +
@.obj_name + '.txt -c -U<xx> -P<xxxxx> -S<xxxxxxxxxx>'
b) How do I execute 1000 or so scripts automatically.
Thanks very much in advance to respond.
Cheers,
SGHi
"subodh97@.gmail.com" wrote:

> Hi,
> We have four different SQL Servers and each server has 2-3 databases.
> The stored procedures on each database use link servers to access each
> other. The link server names are based on the actual server name.
> The Issue:- We are migrating to SQL 2005 and to save on the licence
> cost we are planning to consolidate all our databases on one more
> powerful machine which can take load of peak hours. Now will it be
> advisable to create link servers on the new server by same names as
> existing on the current servers, so that we don't have to change the
> code. Or should we change the code and remove link servers from in the
> code of each SP and trigger.
> I think quires using the link server will be slower then directly
> assessing other databases.
> For Ex <linkserver>.<dbname>.dbo.<tablename> vs. <dbname>..<tablename>
> If the best solution is to change the code then what is the best &
> fast way to do so (there are more then 1000 SP, functions and triggers
> to be changed).
> Following are the few differrent approches I've followed:-
> 1) I tried to change in the syscomments table but SQL Server 2005
> won't allow me to change the sys tables.
> 2) I wrote a SP to find out the search string in the code using
> systables and once I had the object name where the search string is
> used then i imported the code in a temp table using sp_helptext. Then
> removed the linkserver in the temp table and bcp the temp table into a
> txt file (upto this point I have created a script of SP/trigger with
> the changed code)
> The issue:- a) BCP is inserting multiple blank lines in the txt file.
> How do I make sure these unwanted blank lines do not appear in the txt
> file. The reason why I'm more worried abt the blank lines is because
> each SP might go thru multiple change cycles as I described above (if
> different linkservers are used in the same SP/trigger) and every time
> there will be balnak lines added which will blot the code in terms of
> LOC and makes it difficult to maintain long term (readability issues).
> The bcp command I'm using as follows:-
> 'bcp '+ @.db_name + '.dbo.temp_helptext out '+ @.bcp_path + '' +
> @.obj_name + '.txt -c -U<xx> -P<xxxxx> -S<xxxxxxxxxx>'
> b) How do I execute 1000 or so scripts automatically.
> Thanks very much in advance to respond.
> Cheers,
> SG
>
The easiest way would be to script out your stored procedures using the
scripting options in SSMS or Enterprise Manager and then globally replace
each instance with a text editor. If you used source code control for the
your database code this would not be an issue and you can track and control
the changes to the source when you start to implement SQL 2005 features.
John|||SG,
I'm not sure about SQL Server 2005, but with SQL Server 2K I remember having
performance overhead when using a 4 part name with a database on the same
server. SQL Server still goes through unnecessary additional security,
collation conversion, etc. layers as if it was linking to another server.
You might want to consider the following low tech approach:
1. Script out all stored procs to a file.
2. Replace "ServerName." with "".
3. Replace "Create Procedure" with "Alter Procedure".
4. Apply the script.
-- Bill
<subodh97@.gmail.com> wrote in message
news:1173977536.347970.223200@.p15g2000hsd.googlegroups.com...
> Hi,
> We have four different SQL Servers and each server has 2-3 databases.
> The stored procedures on each database use link servers to access each
> other. The link server names are based on the actual server name.
> The Issue:- We are migrating to SQL 2005 and to save on the licence
> cost we are planning to consolidate all our databases on one more
> powerful machine which can take load of peak hours. Now will it be
> advisable to create link servers on the new server by same names as
> existing on the current servers, so that we don't have to change the
> code. Or should we change the code and remove link servers from in the
> code of each SP and trigger.
> I think quires using the link server will be slower then directly
> assessing other databases.
> For Ex <linkserver>.<dbname>.dbo.<tablename> vs. <dbname>..<tablename>
> If the best solution is to change the code then what is the best &
> fast way to do so (there are more then 1000 SP, functions and triggers
> to be changed).
> Following are the few differrent approches I've followed:-
> 1) I tried to change in the syscomments table but SQL Server 2005
> won't allow me to change the sys tables.
> 2) I wrote a SP to find out the search string in the code using
> systables and once I had the object name where the search string is
> used then i imported the code in a temp table using sp_helptext. Then
> removed the linkserver in the temp table and bcp the temp table into a
> txt file (upto this point I have created a script of SP/trigger with
> the changed code)
> The issue:- a) BCP is inserting multiple blank lines in the txt file.
> How do I make sure these unwanted blank lines do not appear in the txt
> file. The reason why I'm more worried abt the blank lines is because
> each SP might go thru multiple change cycles as I described above (if
> different linkservers are used in the same SP/trigger) and every time
> there will be balnak lines added which will blot the code in terms of
> LOC and makes it difficult to maintain long term (readability issues).
> The bcp command I'm using as follows:-
> 'bcp '+ @.db_name + '.dbo.temp_helptext out '+ @.bcp_path + '' +
> @.obj_name + '.txt -c -U<xx> -P<xxxxx> -S<xxxxxxxxxx>'
> b) How do I execute 1000 or so scripts automatically.
> Thanks very much in advance to respond.
> Cheers,
> SG
>|||On Mar 15, 1:39 pm, John Bell <jbellnewspo...@.hotmail.com> wrote:
> Hi
>
> "subod...@.gmail.com" wrote:
>
>
>
>
>
>
>
>
>
>
>
>
>
> The easiest way would be to script out your stored procedures using the
> scripting options in SSMS or Enterprise Manager and then globally replace
> each instance with a text editor. If you used source code control for the
> your database code this would not be an issue and you can track and contro
l
> the changes to the source when you start to implement SQL 2005 features.
> John
Thanks John, The thing is we don't have any source control at this
point (may be future we'll have VSS). And also to run the scripts
manually which I don't want to do on the final day of migration. If
somebody can help me telling the way how to get away with the numerous
blank lines while bcp..ing the code to a text file then half of the
battle is won. Following is the actual code of the SP:-
/*
Author - SG
*/
ALTER PROC [dbo].[sp_use_linkserver]
AS
BEGIN
DECLARE
@.count INT
-- Just a comment
SET @.count = (SELECT count(*) FROM LINKSRV2.testdb.dbo.testtable)
print 'Total # of sites in R3 = ' + CAST(@.count AS VARCHAR(10))
END
Following is the code in txt file generated:-
/*
Author - Subodh Goyal
*/
CREATE PROC [dbo].[sp_use_linkserver]
AS
BEGIN
DECLARE
@.count INT
-- Just a comment
SET @.count = (SELECT count(*) FROM LINKSRV1.testdb.dbo.testtable)
print 'Total # of sites in R3 = ' + CAST(@.count AS VARCHAR(10))
END

Hope this help in understanding the issue.
SG|||On Mar 15, 2:33 pm, "AlterEgo" <altereg...@.dslextreme.com> wrote:[vbcol=seagreen]
> SG,
> I'm not sure about SQL Server 2005, but with SQL Server 2K I remember havi
ng
> performance overhead when using a 4 part name with a database on the same
> server. SQL Server still goes through unnecessary additional security,
> collation conversion, etc. layers as if it was linking to another server.
> You might want to consider the following low tech approach:
> 1. Script out all stored procs to a file.
> 2. Replace "ServerName." with "".
> 3. Replace "Create Procedure" with "Alter Procedure".
> 4. Apply the script.
> -- Bill
> <subod...@.gmail.com> wrote in message
> news:1173977536.347970.223200@.p15g2000hsd.googlegroups.com...
>
>
>
>
>
>
>
>
>
>
>
>
>
>
Thanks Bill,
The only issue in following this approach is sheer amount of work on
the final migration day. There are about 12 different databases which
will be migrating on single machine. Also I can not script all the
databases before hand they need to be scripted on the same day. Also
there are SP. functions and triggers which will make this exercise not
smooth. Anyways this is going to be my last option if nothing else can
be done before the final migration day.
SG|||Hi
sp_helptext puts in the extra lines, you could replace the extra char(10)
and/or char(13) characters using the SQL REPLACE function, but I am not sure
that the method you are using is going to work as it may not reliably produc
e
line breaks in the correct place and you may get words broken.
If you do not want to use scripting then look at using DMO to get the
procedure definition and change them accordingly. This will allow you to onl
y
change the ones where you find the linked servers.
John
"subodh97@.gmail.com" wrote:

> On Mar 15, 1:39 pm, John Bell <jbellnewspo...@.hotmail.com> wrote:
> Thanks John, The thing is we don't have any source control at this
> point (may be future we'll have VSS). And also to run the scripts
> manually which I don't want to do on the final day of migration. If
> somebody can help me telling the way how to get away with the numerous
> blank lines while bcp..ing the code to a text file then half of the
> battle is won. Following is the actual code of the SP:-
> /*
> Author - SG
> */
> ALTER PROC [dbo].[sp_use_linkserver]
> AS
> BEGIN
> DECLARE
> @.count INT
> -- Just a comment
> SET @.count = (SELECT count(*) FROM LINKSRV2.testdb.dbo.testtable)
> print 'Total # of sites in R3 = ' + CAST(@.count AS VARCHAR(10))
> END
>
> Following is the code in txt file generated:-
>
> /*
> Author - Subodh Goyal
> */
>
> CREATE PROC [dbo].[sp_use_linkserver]
>
> AS
>
> BEGIN
>
> DECLARE
> @.count INT
> -- Just a comment
> SET @.count = (SELECT count(*) FROM LINKSRV1.testdb.dbo.testtable)
> print 'Total # of sites in R3 = ' + CAST(@.count AS VARCHAR(10))
> END
>
>
>
>
>
>
>
> Hope this help in understanding the issue.
> SG
>|||Hi
I am surprised that your stored procedures/functions/triggers are so
volatile you can not do this till the last minute? It is not a good idea to
combine an upgrade with major changes to the software unless it is necessary
to make it work on the new system. It is not a good idea to implement these
changes blindly without doing an test migration and regression test cycle
before you implement this on a live system.
Have you run the SQL 2005 upgrade advisor on these systems?
John|||On Mar 15, 9:52 pm, subod...@.gmail.com wrote:
> Hi,
> We have four different SQL Servers and each server has 2-3 databases.
> The stored procedures on each database use link servers to access each
> other. The link server names are based on the actual server name.
> The Issue:- We are migrating to SQL 2005 and to save on the licence
> cost we are planning to consolidate all our databases on one more
> powerful machine which can take load of peak hours. Now will it be
> advisable to create link servers on the new server by same names as
> existing on the current servers, so that we don't have to change the
> code. Or should we change the code and remove link servers from in the
> code of each SP and trigger.
> I think quires using the link server will be slower then directly
> assessing other databases.
> For Ex <linkserver>.<dbname>.dbo.<tablename> vs. <dbname>..<tablename>
> If the best solution is to change the code then what is the best &
> fast way to do so (there are more then 1000 SP, functions and triggers
> to be changed).
> Following are the few differrent approches I've followed:-
> 1) I tried to change in the syscomments table but SQL Server 2005
> won't allow me to change the sys tables.
> 2) I wrote a SP to find out the search string in the code using
> systables and once I had the object name where the search string is
> used then i imported the code in a temp table using sp_helptext. Then
> removed the linkserver in the temp table and bcp the temp table into a
> txt file (upto this point I have created a script of SP/trigger with
> the changed code)
> The issue:- a) BCP is inserting multiple blank lines in the txt file.
> How do I make sure these unwanted blank lines do not appear in the txt
> file. The reason why I'm more worried abt the blank lines is because
> each SP might go thru multiple change cycles as I described above (if
> different linkservers are used in the same SP/trigger) and every time
> there will be balnak lines added which will blot the code in terms of
> LOC and makes it difficult to maintain long term (readability issues).
> The bcp command I'm using as follows:-
> 'bcp '+ @.db_name + '.dbo.temp_helptext out '+ @.bcp_path + '' +
> @.obj_name + '.txt -c -U<xx> -P<xxxxx> -S<xxxxxxxxxx>'
> b) How do I execute 1000 or so scripts automatically.
> Thanks very much in advance to respond.
> Cheers,
> SG
Please check synonym in BOL|||On Mar 16, 4:56 am, John Bell <jbellnewspo...@.hotmail.com> wrote:
> Hi
> I am surprised that your stored procedures/functions/triggers are so
> volatile you can not do this till the last minute? It is not a good idea t
o
> combine an upgrade with major changes to the software unless it is necessa
ry
> to make it work on the new system. It is not a good idea to implement thes
e
> changes blindly without doing an test migration and regression test cycle
> before you implement this on a live system.
> Have you run the SQL 2005 upgrade advisor on these systems?
> John
Hi John, Thanks for the reply. We are getting the new hardware for the
migration and planning to do a mock migration first on the new server
and put that new server and applications on testing mode. Our current
production servers will still run in parallel. There will be a
difference of abt a month or so in mock and actual migration and I
cannot stop any release of new code to the current production servers
during that one month or so period due to the nature of business and
4-5 different teams involved in it. The solution I'm looking is to
generate the right scripts dynamically so that I can incorporate any
changes done during the month of testing in the final migration.
By the way I've run Upgrade adviser on all the servers and there were
no major issues other then few of the DTS packages we have to re-write
in 2005 SSIS.
I'll try DMO option to see how effective and easy it can be.
Thanks,
SG|||On Mar 16, 2:07 pm, subod...@.gmail.com wrote:
> On Mar 16, 4:56 am, John Bell <jbellnewspo...@.hotmail.com> wrote:
> Hi John, Thanks for the reply. We are getting the new hardware for the
> migration and planning to do a mock migration first on the new server
> and put that new server and applications on testing mode. Our current
> production servers will still run in parallel. There will be a
> difference of abt a month or so in mock and actual migration and I
> cannot stop any release of new code to the current production servers
> during that one month or so period due to the nature of business and
> 4-5 different teams involved in it. The solution I'm looking is to
> generate the right scripts dynamically so that I can incorporate any
> changes done during the month of testing in the final migration.
>
Not freezing development is making this task significantly more risky.
Source code control would have made your task significantly easier. An
option may be to have a second cut over for a shorter period where you
run through all the tests and freeze the code for that shorter time? A
different approach would be to take backups of the databases that were
cut over and then compare a restored backup against the live database
at the final cutover using a tool such as Red Gate SQL Compare or
DBGhost. This will show you everything that you will need to change in
the migration. Having two copies of a given database may require
significant disc space, but they do not have to be on the same
machines.

> By the way I've run Upgrade adviser on all the servers and there were
> no major issues other then few of the DTS packages we have to re-write
> in 2005 SSIS.
> I'll try DMO option to see how effective and easy it can be.
> Thanks,
> SG
John

Tuesday, March 27, 2012

changing the server.database against which a script is executed

Hi there:
I write scripts which I must execute in multiple environments (Dev,
Test, UAT, Prod). Dev and Test are located on the same server, so
it's easy to change the db with the USE command. UAT and Prod are on
another server. If I close the script and re-open it, I'm not
prompted for connection information -- it simply defaults to whichever
connection I used last (because I haven't closed SSMS?).
So far, I've use the Object Browser to navigate to an object within
the server.db I want, click New Query, then copy & paste the query
into the new window. This seems pretty cheesy. Is there a more
elegant approach?
Thanks;
Duncan
Duncan A. McRae (google.com@.mcrae.ca) writes:
> I write scripts which I must execute in multiple environments (Dev,
> Test, UAT, Prod). Dev and Test are located on the same server, so
> it's easy to change the db with the USE command. UAT and Prod are on
> another server. If I close the script and re-open it, I'm not
> prompted for connection information -- it simply defaults to whichever
> connection I used last (because I haven't closed SSMS?).
> So far, I've use the Object Browser to navigate to an object within
> the server.db I want, click New Query, then copy & paste the query
> into the new window. This seems pretty cheesy. Is there a more
> elegant approach?
Yes. Right-click in the query window and select Change Connection from
the context menu.
Overall, I prefer Query Analyzer over Mgmt Studio, but being able to
change the connection for a query window is a great feature.
Erland Sommarskog, SQL Server MVP, esquel@.sommarskog.se
Books Online for SQL Server 2005 at
http://www.microsoft.com/technet/prodtechnol/sql/2005/downloads/books.mspx
Books Online for SQL Server 2000 at
http://www.microsoft.com/sql/prodinfo/previousversions/books.mspx

Sunday, March 25, 2012

changing the default port on a clustered 2005 server.

I am unclear as to which Ip address to change the standard port on. When setting up a cluster you have multiple IP address's including the nic that is used for the heartbeat. I have not found any documention on how to change both nodes to listen on a port other than 1433..

Any help would be greatly appreciated.

Set the one titled "IPAll" (you may need to scroll to the bottom of the list). The clustered SQL Server instance will apply it to the actual IP addresses of the cluster.

changing the default port on a clustered 2005 server.

Any help would be greatly appreciated.

Set the one titled "IPAll" (you may need to scroll to the bottom of the list). The clustered SQL Server instance will apply it to the actual IP addresses of the cluster.

Monday, March 19, 2012

Changing SQL Server instance name

Is it possible to change the name of a SQL Server instance on a server running multiple server instances? Or must the server be reinstalled?

Rgds

Bertrand

Hi,

Unfortunatelly, I do not know any solution for renaming instance of SQL server. It is not possible. Try to reinstall with the new instance name of your sql server.

Regards,

Janos

|||There is NO way to do this.

Jens K. Suessmeyer.

http://www.sqlserver2005.de

Wednesday, March 7, 2012

Changing Parameter in SQLDataSource for Multiple DropDowns

I have one page, one connection, and three dropdowns.
The connection looks like (as an example):

<asp:SqlDataSource ID="DropDownConn" runat="server" ConnectionString="<%$ ConnectionStrings:aousConnectionString %>"
SelectCommand="SELECT [Value], [Text] FROM [DropDown] WHERE (([Group] = @.Group) AND ([Viewable] = @.Viewable))">
<SelectParameters>
<asp:Parameter Name="Group" Type="String" />
<asp:Parameter DefaultValue="True" Name="Viewable" Type="Boolean" />
</SelectParameters>
</asp:SqlDataSource>

The DropDowns Look like this:

<asp:DropDownList ID="DropDown1" runat="server"></asp:DropDownList>
<asp:DropDownList ID="DropDown2" runat="server"></asp:DropDownList>
<asp:DropDownList ID="DropDown3" runat="server"></asp:DropDownList
The C# Code I am trying is like this:
DropDownConn.SelectParameters["Group"].Equals("DropDown1");
DropDownConn.SelectParameters["Viewable"].Equals(true);
DropDown1.DataSourceID = "DropDownConn";
DropDown1.DataTextField = "Text";
DropDown1.DataValueField = "Value";
DropDown1.DataBind();

As an example. I can not get it done so that I don't have to create 3 dataconnections. Any help, PLEASE?

It should be done this way..

DropDownConn.SelectParameters["Group"].DefaultValue = "some group value"; // change it accordingly on what data should be displayed in the dropdownlists
DropDownConn.SelectParameters["Viewable"].DefaultValue = "True";

DropDown1.DataSource = DropDownConn;
DropDown1.DataTextField = "Text";
DropDown1.DataValueField = "Value";
DropDown1.DataBind();

// for second

DropDownConn.SelectParameters["Group"].DefaultValue = "some group value for second dropdown"; // change it accordingly on what data should be displayed in the dropdownlists
DropDownConn.SelectParameters["Viewable"].DefaultValue = "True";

DropDown2.DataSource = DropDownConn;
DropDown2.DataTextField = "Text";
DropDown2.DataValueField = "Value";
DropDown2.DataBind();

-----

You can also, refactor the code to set the datasource for the dropdownlist.

Thanks

-Mark post(s) as "Answer" that helped you

|||

Thank you for the answer, and that method does work. Is that really the way we are suppose to do it. The default value is such a poor naming scheme and Microsoft had done such a good job with most other things. Thanks for the the help.

Tuesday, February 14, 2012

Changing default listening port

Hi,
I need to change the default listening port on sql server from 1433 to
another port. The server is being connected to by multiple web sites some
using DSN connections and others not and also via a number of connections
using Enterprise Manager. I'm worried that if I change the port, the
websites will not be able to connect. Can anyone give me some advice on what
will need to be changed to make this update smooth.
Many thanks,
Luke.You can create an alias on the client using "Client Network Utility" and there specify the port
number. Or, you can specify the port number in the server name/connection string. From memory, you
do like SERVERNAME;portnumber, I.e.:
MYSERVER;1067
--
Tibor Karaszi, SQL Server MVP
Archive at: http://groups.google.com/groups?oi=djq&as_ugroup=microsoft.public.sqlserver
"Luke Fields" <lfields228@.hotmaik.com> wrote in message
news:egS7UeluDHA.2520@.TK2MSFTNGP10.phx.gbl...
> Hi,
> I need to change the default listening port on sql server from 1433 to
> another port. The server is being connected to by multiple web sites some
> using DSN connections and others not and also via a number of connections
> using Enterprise Manager. I'm worried that if I change the port, the
> websites will not be able to connect. Can anyone give me some advice on what
> will need to be changed to make this update smooth.
> Many thanks,
> Luke.
>|||You can change the port SQL Server is listening on use the Server Network
Utility in the SQL Server program group. If there are firewalls between the
server and the clients you will need to open the new port on the firewall
so the traffic can get through.
Also if you are using dynamic discovery of the SQL Server on the network
UDP port 1434 will need to be opened as well.
Rand
This posting is provided "as is" with no warranties and confers no rights.

Changing default listening port

quote:

> Hi,
> I need to change the default listening port on sql server from 1433 to
> another port. The server is being connected to by multiple web sites some
> using DSN connections and others not and also via a number of connections
> using Enterprise Manager. I'm worried that if I change the port, the
> websites will not be able to connect. Can anyone give me some advice on wh
at
> will need to be changed to make this update smooth.
> Many thanks,
> Luke.
>

Friday, February 10, 2012

Changing current database during local tx or xa tx

What are the consequences of changing the current database (using "use dbnam
e")
during a transaction?
If I modify multiple databases (on the same Sql Server instance) during a
transaction, will all changes be done in the same transaction (and commited/
rollbacked
atomically)?
Does using XA tx change the answer to the question?Taras Tielkes (taras.tielkes@.gmail.com) writes:
> What are the consequences of changing the current database (using "use
> dbname") during a transaction?
> If I modify multiple databases (on the same Sql Server instance) during
> a transaction, will all changes be done in the same transaction (and
> commited/rollbacked atomically)?
Yes.

> Does using XA tx change the answer to the question?
XA?
Erland Sommarskog, SQL Server MVP, esquel@.sommarskog.se
Books Online for SQL Server SP3 at
http://www.microsoft.com/sql/techin.../2000/books.asp|||Taras Tielkes wrote:
> What are the consequences of changing the current database (using
> "use dbname") during a transaction?
> If I modify multiple databases (on the same Sql Server instance)
> during a transaction, will all changes be done in the same
> transaction (and commited/rollbacked atomically)?
> Does using XA tx change the answer to the question?
Yes. Same transaction. But you cannot use USE statements in stored
procedures or triggers. There's a better way, however. USE, which
physically changes a user's context, is an expensive operation. You're
just better off fully-qualifying the objects you are modifying. Then you
don't have to switch databases:
For example:
Update MyDB.dbo.MyTable SET ...
David Gugick
Quest Software
www.imceda.com
www.quest.com

Thursday, March 29, 2012

Changing the text of the code of multiple stored procedure

Changing the text of the code of multiple stored procedure

Changing the text of the code of multiple stored procedure

Tuesday, March 27, 2012

changing the server.database against which a script is executed

Sunday, March 25, 2012

changing the default port on a clustered 2005 server.

changing the default port on a clustered 2005 server.

Monday, March 19, 2012

Changing SQL Server instance name

Wednesday, March 7, 2012

Changing Parameter in SQLDataSource for Multiple DropDowns

Tuesday, February 14, 2012

Changing default listening port

Changing default listening port

Friday, February 10, 2012

Changing current database during local tx or xa tx

database collation

Blog Archive

About Me