I am modifying a stored procedure to perform a check before updating the
table. I wanted to run tnis if else stmt to see if there is a better way to
perform the task.
The original sp generates a account number for a customer and updates the
table with the new account number.
I want to add a IF stmt to check if the account number already exists before
generating the account number, Print a message that the account # already
exsits and it will keep checking until a new account# appears.
The code:
IF (SELECT @.accountnumber <> @.accountnumberinput) FROM tablekeys
WHERE keyname = 'accountnumber')
BEGIN
UPDATE tablekeys SET @.accountnumberinput = currentvalue = (currentvalue /
2) +
((currentvalue % 2 + ((currentvalue / 8) % 2)) % 2) * POWER(2, 30)
WHERE keyname = 'accountnumber'
SELECT @.accountnumber = db.rcutil_inttobasex(@.accountnumberinput,
'0123456789BCDFGHJKLMNPQRSTVWXYZ')
END
ELSE
PRINT 'Account Number already exists, trying again'Am Mon, 5 Jun 2006 09:03:02 -0700 schrieb SAM:
> I am modifying a stored procedure to perform a check before updating the
> table. I wanted to run tnis if else stmt to see if there is a better way t
o
> perform the task.
> The original sp generates a account number for a customer and updates the
> table with the new account number.
> I want to add a IF stmt to check if the account number already exists befo
re
> generating the account number, Print a message that the account # already
> exsits and it will keep checking until a new account# appears.
>
This is not possible with a stored procedure because a stored proc is not
made for interactive communication. The stored proc can only send back
something (a value, a result set, an error message) to the calling
application, the rest must be done by the application and not by the stored
procedure.
So you can do the check and if it fails you can send back the message to
the calling application, then the calling application shows the message to
the user, the user enters a new account number and the application calls
the stored proc again with the new account number and so on...
By the way, your IF looks wrong for me, is this working? I think not.
For example, if i want to check if the @.newaccountnumber exists in the
accounttable, i would write the statement this way:
IF exists(select * from accounttable where accountnumber =
@.newaccountnumber) begin
raiserror('account number exists',16,1)
return -1
end
...
But if i am wrong, forget this sample :-))
bye, Helmut|||The user doesn't enter the account number. The stored procedure generates th
e
account number and assigns it the customer or user by updating the table.
I was just displaying a message but it is not necessary. I just wanted to
perform a check within the procedure to check the account prior to generatin
g
the new account number.
Therefore, there is nothing entered by the user or application for
interaction with the sp.
Would I still use your sample?
"Helmut Woess" wrote:
> Am Mon, 5 Jun 2006 09:03:02 -0700 schrieb SAM:
>
> This is not possible with a stored procedure because a stored proc is not
> made for interactive communication. The stored proc can only send back
> something (a value, a result set, an error message) to the calling
> application, the rest must be done by the application and not by the store
d
> procedure.
> So you can do the check and if it fails you can send back the message to
> the calling application, then the calling application shows the message to
> the user, the user enters a new account number and the application calls
> the stored proc again with the new account number and so on...
> By the way, your IF looks wrong for me, is this working? I think not.
> For example, if i want to check if the @.newaccountnumber exists in the
> accounttable, i would write the statement this way:
> IF exists(select * from accounttable where accountnumber =
> @.newaccountnumber) begin
> raiserror('account number exists',16,1)
> return -1
> end
> ...
> But if i am wrong, forget this sample :-))
> bye, Helmut
>|||Am Mon, 5 Jun 2006 09:55:01 -0700 schrieb SAM:
> The user doesn't enter the account number. The stored procedure generates
the
> account number and assigns it the customer or user by updating the table.
> I was just displaying a message but it is not necessary. I just wanted to
> perform a check within the procedure to check the account prior to generat
ing
> the new account number.
> Therefore, there is nothing entered by the user or application for
> interaction with the sp.
> Would I still use your sample?
Hm, okay, sorry, my english is not the best, propably i missunderstand you.
And i cannot find out, why there is @.accountnumberinput, if nothing is
entered by user or application. So i don't know how you will generate a
unique accountnumber if the first generated number is not unique..?
I would need more input because i don't understand your question :-(
bye, Helmut|||It is grapping that value from another table.
When a new user is added via the Web UI, a new account number generated and
added to the table along with the customer information that was entered by
the user. The account number is not entered by the user, the system assigns
this number via the store procedure.
Currently, when a new user is added under an exisitng account, that user
shares the same account number. We do not want this to happen. We want each
user, rather with the same company or under the same account name to have
their own account number.
Therefore, I wanted to alter the existing stored procedure to add a check
clause. If the new user that is being added and the system tries to assigned
an existing account number to the new user, I want a flag or check clause to
not assigned the user the same acct # but generated a new one and assigned i
t
to the new user.
I hope that makes more sense.
Actually, I think I need to perform this check in another stored procedure
that is creating the account information. I will post that code in a few
minutes. Thanks
"Helmut Woess" wrote:
> Am Mon, 5 Jun 2006 09:55:01 -0700 schrieb SAM:
>
> Hm, okay, sorry, my english is not the best, propably i missunderstand you
.
> And i cannot find out, why there is @.accountnumberinput, if nothing is
> entered by user or application. So i don't know how you will generate a
> unique accountnumber if the first generated number is not unique..?
> I would need more input because i don't understand your question :-(
> bye, Helmut
>sql
Showing posts with label modifying. Show all posts
Showing posts with label modifying. Show all posts
Tuesday, March 27, 2012
Thursday, March 8, 2012
Alter Index issue & Try/Catch question
All,
I'm currently modifying a BOL procedure to rebuild/reorganize my
indexes. I've changed the ALTER INDEX command so that it performs the
rebuild online. The first time I ran the script, it errored with the
following error below:
Msg 2725, Level 16, State 2, Line 1
Online index operation cannot be performed for index
'Company$Attachment$0' because the index contains column 'Entry Pointer
ID' of data type text, ntext, image, varchar(max), nvarchar(max),
varbinary(max) or xml. For non-clustered index the column could be an
include column of the index, for clustered index it could be any column
of the table. In case of drop_existing the column could be part of new
or old index. The operation must be performed offline.
At this point, I have tried to integrate the TRY/CATCH routine so that
when this error appears, the script executes the ALTER INDEXES offline
instead (script below). I was wondering if there is a way to possible
write out the error to a log file perhaps?
Thanks,
Ian
SET NOCOUNT ON;
DECLARE @.objectid int;
DECLARE @.indexid int;
DECLARE @.partitioncount bigint;
DECLARE @.schemaname sysname;
DECLARE @.objectname sysname;
DECLARE @.indexname sysname;
DECLARE @.partitionnum bigint;
DECLARE @.partitions bigint;
DECLARE @.frag float;
DECLARE @.command varchar(8000);
-- ensure the temporary table does not exist
IF EXISTS (SELECT name FROM sys.objects WHERE name = 'work_to_do')
DROP TABLE work_to_do;
-- conditionally select from the function, converting object and index
IDs to names.
SELECT
object_id AS objectid,
index_id AS indexid,
partition_number as partitionnum,
avg_fragmentation_in_percent as frag
INTO work_to_do
FROM sys.dm_db_index_physical_stats (5, NULL, NULL , NULL, 'LIMITED')
WHERE avg_fragmentation_in_percent > 10.0 AND index_id > 0;
-- Declare the cursor for the list of partitions to be processed.
DECLARE partitions CURSOR FOR SELECT * FROM work_to_do;
-- Open the cursor.
OPEN partitions;
-- Loop through the partitions.
FETCH NEXT
FROM partitions
INTO @.objectid, @.indexid, @.partitionnum, @.frag;
WHILE @.@.FETCH_STATUS = 0
BEGIN;
SELECT @.objectname = o.name, @.schemaname = s.name
FROM sys.objects AS o
JOIN sys.schemas as s ON s.schema_id = o.schema_id
WHERE o.object_id = @.objectid;
SELECT @.indexname = name
FROM sys.indexes
WHERE object_id = @.objectid AND index_id = @.indexid;
SELECT @.partitioncount = count (*)
FROM sys.partitions
WHERE object_id = @.objectid AND index_id = @.indexid;
-- 30 is an arbitrary decision point at which to switch between
reorganizing and rebuilding
IF @.frag < 30.0
BEGIN;
SELECT @.command = 'ALTER INDEX [' + @.indexname + '] ON ' + '[' +
@.objectname + '] REORGANIZE';
IF @.partitioncount > 1
SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
@.partitionnum);
PRINT (@.command);
EXEC (@.command);
END;
IF @.frag >= 30.0
BEGIN;
SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
@.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
STATISTICS_NORECOMPUTE=OFF) ';
IF @.partitioncount > 1
SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
@.partitionnum);
BEGIN TRY
EXEC (@.command);
END TRY
BEGIN CATCH
SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
@.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
STATISTICS_NORECOMPUTE=OFF) ';
EXEC (@.command);
END CATCH
PRINT (@.command);
END;
PRINT 'Executed ' + @.command;
PRINT
'----';
FETCH NEXT FROM partitions INTO @.objectid, @.indexid, @.partitionnum,
@.frag;
END;
-- Close and deallocate the cursor.
CLOSE partitions;
DEALLOCATE partitions;
To write a event log entry use: RAISERROR
RAISERROR ('Something happened and needs to be logged', 10, 1 ) WITH LOG
Arnie Rowland, Ph.D.
Westwood Consulting, Inc
Most good judgment comes from experience.
Most experience comes from bad judgment.
- Anonymous
You can't help someone get up a hill without getting a little closer to the
top yourself.
- H. Norman Schwarzkopf
<theredmiata@.hotmail.com> wrote in message
news:1165272932.638685.169330@.n67g2000cwd.googlegr oups.com...
> All,
> I'm currently modifying a BOL procedure to rebuild/reorganize my
> indexes. I've changed the ALTER INDEX command so that it performs the
> rebuild online. The first time I ran the script, it errored with the
> following error below:
> Msg 2725, Level 16, State 2, Line 1
> Online index operation cannot be performed for index
> 'Company$Attachment$0' because the index contains column 'Entry Pointer
> ID' of data type text, ntext, image, varchar(max), nvarchar(max),
> varbinary(max) or xml. For non-clustered index the column could be an
> include column of the index, for clustered index it could be any column
> of the table. In case of drop_existing the column could be part of new
> or old index. The operation must be performed offline.
> At this point, I have tried to integrate the TRY/CATCH routine so that
> when this error appears, the script executes the ALTER INDEXES offline
> instead (script below). I was wondering if there is a way to possible
> write out the error to a log file perhaps?
> Thanks,
> Ian
>
> SET NOCOUNT ON;
> DECLARE @.objectid int;
> DECLARE @.indexid int;
> DECLARE @.partitioncount bigint;
> DECLARE @.schemaname sysname;
> DECLARE @.objectname sysname;
> DECLARE @.indexname sysname;
> DECLARE @.partitionnum bigint;
> DECLARE @.partitions bigint;
> DECLARE @.frag float;
> DECLARE @.command varchar(8000);
> -- ensure the temporary table does not exist
> IF EXISTS (SELECT name FROM sys.objects WHERE name = 'work_to_do')
> DROP TABLE work_to_do;
> -- conditionally select from the function, converting object and index
> IDs to names.
> SELECT
> object_id AS objectid,
> index_id AS indexid,
> partition_number as partitionnum,
> avg_fragmentation_in_percent as frag
> INTO work_to_do
> FROM sys.dm_db_index_physical_stats (5, NULL, NULL , NULL, 'LIMITED')
> WHERE avg_fragmentation_in_percent > 10.0 AND index_id > 0;
> -- Declare the cursor for the list of partitions to be processed.
> DECLARE partitions CURSOR FOR SELECT * FROM work_to_do;
> -- Open the cursor.
> OPEN partitions;
> -- Loop through the partitions.
> FETCH NEXT
> FROM partitions
> INTO @.objectid, @.indexid, @.partitionnum, @.frag;
> WHILE @.@.FETCH_STATUS = 0
> BEGIN;
> SELECT @.objectname = o.name, @.schemaname = s.name
> FROM sys.objects AS o
> JOIN sys.schemas as s ON s.schema_id = o.schema_id
> WHERE o.object_id = @.objectid;
> SELECT @.indexname = name
> FROM sys.indexes
> WHERE object_id = @.objectid AND index_id = @.indexid;
> SELECT @.partitioncount = count (*)
> FROM sys.partitions
> WHERE object_id = @.objectid AND index_id = @.indexid;
> -- 30 is an arbitrary decision point at which to switch between
> reorganizing and rebuilding
> IF @.frag < 30.0
> BEGIN;
> SELECT @.command = 'ALTER INDEX [' + @.indexname + '] ON ' + '[' +
> @.objectname + '] REORGANIZE';
> IF @.partitioncount > 1
> SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
> @.partitionnum);
> PRINT (@.command);
> EXEC (@.command);
> END;
> IF @.frag >= 30.0
> BEGIN;
> SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
> @.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
> STATISTICS_NORECOMPUTE=OFF) ';
> IF @.partitioncount > 1
> SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
> @.partitionnum);
> BEGIN TRY
> EXEC (@.command);
> END TRY
> BEGIN CATCH
> SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
> @.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
> STATISTICS_NORECOMPUTE=OFF) ';
> EXEC (@.command);
> END CATCH
> PRINT (@.command);
> END;
> PRINT 'Executed ' + @.command;
> PRINT
> '----';
> FETCH NEXT FROM partitions INTO @.objectid, @.indexid, @.partitionnum,
> @.frag;
> END;
> -- Close and deallocate the cursor.
> CLOSE partitions;
> DEALLOCATE partitions;
>
I'm currently modifying a BOL procedure to rebuild/reorganize my
indexes. I've changed the ALTER INDEX command so that it performs the
rebuild online. The first time I ran the script, it errored with the
following error below:
Msg 2725, Level 16, State 2, Line 1
Online index operation cannot be performed for index
'Company$Attachment$0' because the index contains column 'Entry Pointer
ID' of data type text, ntext, image, varchar(max), nvarchar(max),
varbinary(max) or xml. For non-clustered index the column could be an
include column of the index, for clustered index it could be any column
of the table. In case of drop_existing the column could be part of new
or old index. The operation must be performed offline.
At this point, I have tried to integrate the TRY/CATCH routine so that
when this error appears, the script executes the ALTER INDEXES offline
instead (script below). I was wondering if there is a way to possible
write out the error to a log file perhaps?
Thanks,
Ian
SET NOCOUNT ON;
DECLARE @.objectid int;
DECLARE @.indexid int;
DECLARE @.partitioncount bigint;
DECLARE @.schemaname sysname;
DECLARE @.objectname sysname;
DECLARE @.indexname sysname;
DECLARE @.partitionnum bigint;
DECLARE @.partitions bigint;
DECLARE @.frag float;
DECLARE @.command varchar(8000);
-- ensure the temporary table does not exist
IF EXISTS (SELECT name FROM sys.objects WHERE name = 'work_to_do')
DROP TABLE work_to_do;
-- conditionally select from the function, converting object and index
IDs to names.
SELECT
object_id AS objectid,
index_id AS indexid,
partition_number as partitionnum,
avg_fragmentation_in_percent as frag
INTO work_to_do
FROM sys.dm_db_index_physical_stats (5, NULL, NULL , NULL, 'LIMITED')
WHERE avg_fragmentation_in_percent > 10.0 AND index_id > 0;
-- Declare the cursor for the list of partitions to be processed.
DECLARE partitions CURSOR FOR SELECT * FROM work_to_do;
-- Open the cursor.
OPEN partitions;
-- Loop through the partitions.
FETCH NEXT
FROM partitions
INTO @.objectid, @.indexid, @.partitionnum, @.frag;
WHILE @.@.FETCH_STATUS = 0
BEGIN;
SELECT @.objectname = o.name, @.schemaname = s.name
FROM sys.objects AS o
JOIN sys.schemas as s ON s.schema_id = o.schema_id
WHERE o.object_id = @.objectid;
SELECT @.indexname = name
FROM sys.indexes
WHERE object_id = @.objectid AND index_id = @.indexid;
SELECT @.partitioncount = count (*)
FROM sys.partitions
WHERE object_id = @.objectid AND index_id = @.indexid;
-- 30 is an arbitrary decision point at which to switch between
reorganizing and rebuilding
IF @.frag < 30.0
BEGIN;
SELECT @.command = 'ALTER INDEX [' + @.indexname + '] ON ' + '[' +
@.objectname + '] REORGANIZE';
IF @.partitioncount > 1
SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
@.partitionnum);
PRINT (@.command);
EXEC (@.command);
END;
IF @.frag >= 30.0
BEGIN;
SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
@.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
STATISTICS_NORECOMPUTE=OFF) ';
IF @.partitioncount > 1
SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
@.partitionnum);
BEGIN TRY
EXEC (@.command);
END TRY
BEGIN CATCH
SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
@.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
STATISTICS_NORECOMPUTE=OFF) ';
EXEC (@.command);
END CATCH
PRINT (@.command);
END;
PRINT 'Executed ' + @.command;
'----';
FETCH NEXT FROM partitions INTO @.objectid, @.indexid, @.partitionnum,
@.frag;
END;
-- Close and deallocate the cursor.
CLOSE partitions;
DEALLOCATE partitions;
To write a event log entry use: RAISERROR
RAISERROR ('Something happened and needs to be logged', 10, 1 ) WITH LOG
Arnie Rowland, Ph.D.
Westwood Consulting, Inc
Most good judgment comes from experience.
Most experience comes from bad judgment.
- Anonymous
You can't help someone get up a hill without getting a little closer to the
top yourself.
- H. Norman Schwarzkopf
<theredmiata@.hotmail.com> wrote in message
news:1165272932.638685.169330@.n67g2000cwd.googlegr oups.com...
> All,
> I'm currently modifying a BOL procedure to rebuild/reorganize my
> indexes. I've changed the ALTER INDEX command so that it performs the
> rebuild online. The first time I ran the script, it errored with the
> following error below:
> Msg 2725, Level 16, State 2, Line 1
> Online index operation cannot be performed for index
> 'Company$Attachment$0' because the index contains column 'Entry Pointer
> ID' of data type text, ntext, image, varchar(max), nvarchar(max),
> varbinary(max) or xml. For non-clustered index the column could be an
> include column of the index, for clustered index it could be any column
> of the table. In case of drop_existing the column could be part of new
> or old index. The operation must be performed offline.
> At this point, I have tried to integrate the TRY/CATCH routine so that
> when this error appears, the script executes the ALTER INDEXES offline
> instead (script below). I was wondering if there is a way to possible
> write out the error to a log file perhaps?
> Thanks,
> Ian
>
> SET NOCOUNT ON;
> DECLARE @.objectid int;
> DECLARE @.indexid int;
> DECLARE @.partitioncount bigint;
> DECLARE @.schemaname sysname;
> DECLARE @.objectname sysname;
> DECLARE @.indexname sysname;
> DECLARE @.partitionnum bigint;
> DECLARE @.partitions bigint;
> DECLARE @.frag float;
> DECLARE @.command varchar(8000);
> -- ensure the temporary table does not exist
> IF EXISTS (SELECT name FROM sys.objects WHERE name = 'work_to_do')
> DROP TABLE work_to_do;
> -- conditionally select from the function, converting object and index
> IDs to names.
> SELECT
> object_id AS objectid,
> index_id AS indexid,
> partition_number as partitionnum,
> avg_fragmentation_in_percent as frag
> INTO work_to_do
> FROM sys.dm_db_index_physical_stats (5, NULL, NULL , NULL, 'LIMITED')
> WHERE avg_fragmentation_in_percent > 10.0 AND index_id > 0;
> -- Declare the cursor for the list of partitions to be processed.
> DECLARE partitions CURSOR FOR SELECT * FROM work_to_do;
> -- Open the cursor.
> OPEN partitions;
> -- Loop through the partitions.
> FETCH NEXT
> FROM partitions
> INTO @.objectid, @.indexid, @.partitionnum, @.frag;
> WHILE @.@.FETCH_STATUS = 0
> BEGIN;
> SELECT @.objectname = o.name, @.schemaname = s.name
> FROM sys.objects AS o
> JOIN sys.schemas as s ON s.schema_id = o.schema_id
> WHERE o.object_id = @.objectid;
> SELECT @.indexname = name
> FROM sys.indexes
> WHERE object_id = @.objectid AND index_id = @.indexid;
> SELECT @.partitioncount = count (*)
> FROM sys.partitions
> WHERE object_id = @.objectid AND index_id = @.indexid;
> -- 30 is an arbitrary decision point at which to switch between
> reorganizing and rebuilding
> IF @.frag < 30.0
> BEGIN;
> SELECT @.command = 'ALTER INDEX [' + @.indexname + '] ON ' + '[' +
> @.objectname + '] REORGANIZE';
> IF @.partitioncount > 1
> SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
> @.partitionnum);
> PRINT (@.command);
> EXEC (@.command);
> END;
> IF @.frag >= 30.0
> BEGIN;
> SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
> @.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
> STATISTICS_NORECOMPUTE=OFF) ';
> IF @.partitioncount > 1
> SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
> @.partitionnum);
> BEGIN TRY
> EXEC (@.command);
> END TRY
> BEGIN CATCH
> SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
> @.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
> STATISTICS_NORECOMPUTE=OFF) ';
> EXEC (@.command);
> END CATCH
> PRINT (@.command);
> END;
> PRINT 'Executed ' + @.command;
> '----';
> FETCH NEXT FROM partitions INTO @.objectid, @.indexid, @.partitionnum,
> @.frag;
> END;
> -- Close and deallocate the cursor.
> CLOSE partitions;
> DEALLOCATE partitions;
>
Alter Index issue & Try/Catch question
All,
I'm currently modifying a BOL procedure to rebuild/reorganize my
indexes. I've changed the ALTER INDEX command so that it performs the
rebuild online. The first time I ran the script, it errored with the
following error below:
Msg 2725, Level 16, State 2, Line 1
Online index operation cannot be performed for index
'Company$Attachment$0' because the index contains column 'Entry Pointer
ID' of data type text, ntext, image, varchar(max), nvarchar(max),
varbinary(max) or xml. For non-clustered index the column could be an
include column of the index, for clustered index it could be any column
of the table. In case of drop_existing the column could be part of new
or old index. The operation must be performed offline.
At this point, I have tried to integrate the TRY/CATCH routine so that
when this error appears, the script executes the ALTER INDEXES offline
instead (script below). I was wondering if there is a way to possible
write out the error to a log file perhaps?
Thanks,
Ian
SET NOCOUNT ON;
DECLARE @.objectid int;
DECLARE @.indexid int;
DECLARE @.partitioncount bigint;
DECLARE @.schemaname sysname;
DECLARE @.objectname sysname;
DECLARE @.indexname sysname;
DECLARE @.partitionnum bigint;
DECLARE @.partitions bigint;
DECLARE @.frag float;
DECLARE @.command varchar(8000);
-- ensure the temporary table does not exist
IF EXISTS (SELECT name FROM sys.objects WHERE name = 'work_to_do')
DROP TABLE work_to_do;
-- conditionally select from the function, converting object and index
IDs to names.
SELECT
object_id AS objectid,
index_id AS indexid,
partition_number as partitionnum,
avg_fragmentation_in_percent as frag
INTO work_to_do
FROM sys.dm_db_index_physical_stats (5, NULL, NULL , NULL, 'LIMITED')
WHERE avg_fragmentation_in_percent > 10.0 AND index_id > 0;
-- Declare the cursor for the list of partitions to be processed.
DECLARE partitions CURSOR FOR SELECT * FROM work_to_do;
-- Open the cursor.
OPEN partitions;
-- Loop through the partitions.
FETCH NEXT
FROM partitions
INTO @.objectid, @.indexid, @.partitionnum, @.frag;
WHILE @.@.FETCH_STATUS = 0
BEGIN;
SELECT @.objectname = o.name, @.schemaname = s.name
FROM sys.objects AS o
JOIN sys.schemas as s ON s.schema_id = o.schema_id
WHERE o.object_id = @.objectid;
SELECT @.indexname = name
FROM sys.indexes
WHERE object_id = @.objectid AND index_id = @.indexid;
SELECT @.partitioncount = count (*)
FROM sys.partitions
WHERE object_id = @.objectid AND index_id = @.indexid;
-- 30 is an arbitrary decision point at which to switch between
reorganizing and rebuilding
IF @.frag < 30.0
BEGIN;
SELECT @.command = 'ALTER INDEX [' + @.indexname + '] ON ' + '[' +
@.objectname + '] REORGANIZE';
IF @.partitioncount > 1
SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
@.partitionnum);
PRINT (@.command);
EXEC (@.command);
END;
IF @.frag >= 30.0
BEGIN;
SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
@.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
STATISTICS_NORECOMPUTE=OFF) ';
IF @.partitioncount > 1
SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
@.partitionnum);
BEGIN TRY
EXEC (@.command);
END TRY
BEGIN CATCH
SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
@.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
STATISTICS_NORECOMPUTE=OFF) ';
EXEC (@.command);
END CATCH
PRINT (@.command);
END;
PRINT 'Executed ' + @.command;
PRINT
'----
--';
FETCH NEXT FROM partitions INTO @.objectid, @.indexid, @.partitionnum,
@.frag;
END;
-- Close and deallocate the cursor.
CLOSE partitions;
DEALLOCATE partitions;To write a event log entry use: RAISERROR
RAISERROR ('Something happened and needs to be logged', 10, 1 ) WITH LOG
Arnie Rowland, Ph.D.
Westwood Consulting, Inc
Most good judgment comes from experience.
Most experience comes from bad judgment.
- Anonymous
You can't help someone get up a hill without getting a little closer to the
top yourself.
- H. Norman Schwarzkopf
<theredmiata@.hotmail.com> wrote in message
news:1165272932.638685.169330@.n67g2000cwd.googlegroups.com...
> All,
> I'm currently modifying a BOL procedure to rebuild/reorganize my
> indexes. I've changed the ALTER INDEX command so that it performs the
> rebuild online. The first time I ran the script, it errored with the
> following error below:
> Msg 2725, Level 16, State 2, Line 1
> Online index operation cannot be performed for index
> 'Company$Attachment$0' because the index contains column 'Entry Pointer
> ID' of data type text, ntext, image, varchar(max), nvarchar(max),
> varbinary(max) or xml. For non-clustered index the column could be an
> include column of the index, for clustered index it could be any column
> of the table. In case of drop_existing the column could be part of new
> or old index. The operation must be performed offline.
> At this point, I have tried to integrate the TRY/CATCH routine so that
> when this error appears, the script executes the ALTER INDEXES offline
> instead (script below). I was wondering if there is a way to possible
> write out the error to a log file perhaps?
> Thanks,
> Ian
>
> SET NOCOUNT ON;
> DECLARE @.objectid int;
> DECLARE @.indexid int;
> DECLARE @.partitioncount bigint;
> DECLARE @.schemaname sysname;
> DECLARE @.objectname sysname;
> DECLARE @.indexname sysname;
> DECLARE @.partitionnum bigint;
> DECLARE @.partitions bigint;
> DECLARE @.frag float;
> DECLARE @.command varchar(8000);
> -- ensure the temporary table does not exist
> IF EXISTS (SELECT name FROM sys.objects WHERE name = 'work_to_do')
> DROP TABLE work_to_do;
> -- conditionally select from the function, converting object and index
> IDs to names.
> SELECT
> object_id AS objectid,
> index_id AS indexid,
> partition_number as partitionnum,
> avg_fragmentation_in_percent as frag
> INTO work_to_do
> FROM sys.dm_db_index_physical_stats (5, NULL, NULL , NULL, 'LIMITED')
> WHERE avg_fragmentation_in_percent > 10.0 AND index_id > 0;
> -- Declare the cursor for the list of partitions to be processed.
> DECLARE partitions CURSOR FOR SELECT * FROM work_to_do;
> -- Open the cursor.
> OPEN partitions;
> -- Loop through the partitions.
> FETCH NEXT
> FROM partitions
> INTO @.objectid, @.indexid, @.partitionnum, @.frag;
> WHILE @.@.FETCH_STATUS = 0
> BEGIN;
> SELECT @.objectname = o.name, @.schemaname = s.name
> FROM sys.objects AS o
> JOIN sys.schemas as s ON s.schema_id = o.schema_id
> WHERE o.object_id = @.objectid;
> SELECT @.indexname = name
> FROM sys.indexes
> WHERE object_id = @.objectid AND index_id = @.indexid;
> SELECT @.partitioncount = count (*)
> FROM sys.partitions
> WHERE object_id = @.objectid AND index_id = @.indexid;
> -- 30 is an arbitrary decision point at which to switch between
> reorganizing and rebuilding
> IF @.frag < 30.0
> BEGIN;
> SELECT @.command = 'ALTER INDEX [' + @.indexname + '] ON ' + '['
+
> @.objectname + '] REORGANIZE';
> IF @.partitioncount > 1
> SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
> @.partitionnum);
> PRINT (@.command);
> EXEC (@.command);
> END;
> IF @.frag >= 30.0
> BEGIN;
> SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
> @.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
> STATISTICS_NORECOMPUTE=OFF) ';
> IF @.partitioncount > 1
> SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
> @.partitionnum);
> BEGIN TRY
> EXEC (@.command);
> END TRY
> BEGIN CATCH
> SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
> @.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
> STATISTICS_NORECOMPUTE=OFF) ';
> EXEC (@.command);
> END CATCH
> PRINT (@.command);
> END;
> PRINT 'Executed ' + @.command;
> PRINT
> '----
--';
> FETCH NEXT FROM partitions INTO @.objectid, @.indexid, @.partitionnum,
> @.frag;
> END;
> -- Close and deallocate the cursor.
> CLOSE partitions;
> DEALLOCATE partitions;
>
I'm currently modifying a BOL procedure to rebuild/reorganize my
indexes. I've changed the ALTER INDEX command so that it performs the
rebuild online. The first time I ran the script, it errored with the
following error below:
Msg 2725, Level 16, State 2, Line 1
Online index operation cannot be performed for index
'Company$Attachment$0' because the index contains column 'Entry Pointer
ID' of data type text, ntext, image, varchar(max), nvarchar(max),
varbinary(max) or xml. For non-clustered index the column could be an
include column of the index, for clustered index it could be any column
of the table. In case of drop_existing the column could be part of new
or old index. The operation must be performed offline.
At this point, I have tried to integrate the TRY/CATCH routine so that
when this error appears, the script executes the ALTER INDEXES offline
instead (script below). I was wondering if there is a way to possible
write out the error to a log file perhaps?
Thanks,
Ian
SET NOCOUNT ON;
DECLARE @.objectid int;
DECLARE @.indexid int;
DECLARE @.partitioncount bigint;
DECLARE @.schemaname sysname;
DECLARE @.objectname sysname;
DECLARE @.indexname sysname;
DECLARE @.partitionnum bigint;
DECLARE @.partitions bigint;
DECLARE @.frag float;
DECLARE @.command varchar(8000);
-- ensure the temporary table does not exist
IF EXISTS (SELECT name FROM sys.objects WHERE name = 'work_to_do')
DROP TABLE work_to_do;
-- conditionally select from the function, converting object and index
IDs to names.
SELECT
object_id AS objectid,
index_id AS indexid,
partition_number as partitionnum,
avg_fragmentation_in_percent as frag
INTO work_to_do
FROM sys.dm_db_index_physical_stats (5, NULL, NULL , NULL, 'LIMITED')
WHERE avg_fragmentation_in_percent > 10.0 AND index_id > 0;
-- Declare the cursor for the list of partitions to be processed.
DECLARE partitions CURSOR FOR SELECT * FROM work_to_do;
-- Open the cursor.
OPEN partitions;
-- Loop through the partitions.
FETCH NEXT
FROM partitions
INTO @.objectid, @.indexid, @.partitionnum, @.frag;
WHILE @.@.FETCH_STATUS = 0
BEGIN;
SELECT @.objectname = o.name, @.schemaname = s.name
FROM sys.objects AS o
JOIN sys.schemas as s ON s.schema_id = o.schema_id
WHERE o.object_id = @.objectid;
SELECT @.indexname = name
FROM sys.indexes
WHERE object_id = @.objectid AND index_id = @.indexid;
SELECT @.partitioncount = count (*)
FROM sys.partitions
WHERE object_id = @.objectid AND index_id = @.indexid;
-- 30 is an arbitrary decision point at which to switch between
reorganizing and rebuilding
IF @.frag < 30.0
BEGIN;
SELECT @.command = 'ALTER INDEX [' + @.indexname + '] ON ' + '[' +
@.objectname + '] REORGANIZE';
IF @.partitioncount > 1
SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
@.partitionnum);
PRINT (@.command);
EXEC (@.command);
END;
IF @.frag >= 30.0
BEGIN;
SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
@.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
STATISTICS_NORECOMPUTE=OFF) ';
IF @.partitioncount > 1
SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
@.partitionnum);
BEGIN TRY
EXEC (@.command);
END TRY
BEGIN CATCH
SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
@.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
STATISTICS_NORECOMPUTE=OFF) ';
EXEC (@.command);
END CATCH
PRINT (@.command);
END;
PRINT 'Executed ' + @.command;
'----
--';
FETCH NEXT FROM partitions INTO @.objectid, @.indexid, @.partitionnum,
@.frag;
END;
-- Close and deallocate the cursor.
CLOSE partitions;
DEALLOCATE partitions;To write a event log entry use: RAISERROR
RAISERROR ('Something happened and needs to be logged', 10, 1 ) WITH LOG
Arnie Rowland, Ph.D.
Westwood Consulting, Inc
Most good judgment comes from experience.
Most experience comes from bad judgment.
- Anonymous
You can't help someone get up a hill without getting a little closer to the
top yourself.
- H. Norman Schwarzkopf
<theredmiata@.hotmail.com> wrote in message
news:1165272932.638685.169330@.n67g2000cwd.googlegroups.com...
> All,
> I'm currently modifying a BOL procedure to rebuild/reorganize my
> indexes. I've changed the ALTER INDEX command so that it performs the
> rebuild online. The first time I ran the script, it errored with the
> following error below:
> Msg 2725, Level 16, State 2, Line 1
> Online index operation cannot be performed for index
> 'Company$Attachment$0' because the index contains column 'Entry Pointer
> ID' of data type text, ntext, image, varchar(max), nvarchar(max),
> varbinary(max) or xml. For non-clustered index the column could be an
> include column of the index, for clustered index it could be any column
> of the table. In case of drop_existing the column could be part of new
> or old index. The operation must be performed offline.
> At this point, I have tried to integrate the TRY/CATCH routine so that
> when this error appears, the script executes the ALTER INDEXES offline
> instead (script below). I was wondering if there is a way to possible
> write out the error to a log file perhaps?
> Thanks,
> Ian
>
> SET NOCOUNT ON;
> DECLARE @.objectid int;
> DECLARE @.indexid int;
> DECLARE @.partitioncount bigint;
> DECLARE @.schemaname sysname;
> DECLARE @.objectname sysname;
> DECLARE @.indexname sysname;
> DECLARE @.partitionnum bigint;
> DECLARE @.partitions bigint;
> DECLARE @.frag float;
> DECLARE @.command varchar(8000);
> -- ensure the temporary table does not exist
> IF EXISTS (SELECT name FROM sys.objects WHERE name = 'work_to_do')
> DROP TABLE work_to_do;
> -- conditionally select from the function, converting object and index
> IDs to names.
> SELECT
> object_id AS objectid,
> index_id AS indexid,
> partition_number as partitionnum,
> avg_fragmentation_in_percent as frag
> INTO work_to_do
> FROM sys.dm_db_index_physical_stats (5, NULL, NULL , NULL, 'LIMITED')
> WHERE avg_fragmentation_in_percent > 10.0 AND index_id > 0;
> -- Declare the cursor for the list of partitions to be processed.
> DECLARE partitions CURSOR FOR SELECT * FROM work_to_do;
> -- Open the cursor.
> OPEN partitions;
> -- Loop through the partitions.
> FETCH NEXT
> FROM partitions
> INTO @.objectid, @.indexid, @.partitionnum, @.frag;
> WHILE @.@.FETCH_STATUS = 0
> BEGIN;
> SELECT @.objectname = o.name, @.schemaname = s.name
> FROM sys.objects AS o
> JOIN sys.schemas as s ON s.schema_id = o.schema_id
> WHERE o.object_id = @.objectid;
> SELECT @.indexname = name
> FROM sys.indexes
> WHERE object_id = @.objectid AND index_id = @.indexid;
> SELECT @.partitioncount = count (*)
> FROM sys.partitions
> WHERE object_id = @.objectid AND index_id = @.indexid;
> -- 30 is an arbitrary decision point at which to switch between
> reorganizing and rebuilding
> IF @.frag < 30.0
> BEGIN;
> SELECT @.command = 'ALTER INDEX [' + @.indexname + '] ON ' + '['
+
> @.objectname + '] REORGANIZE';
> IF @.partitioncount > 1
> SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
> @.partitionnum);
> PRINT (@.command);
> EXEC (@.command);
> END;
> IF @.frag >= 30.0
> BEGIN;
> SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
> @.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
> STATISTICS_NORECOMPUTE=OFF) ';
> IF @.partitioncount > 1
> SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
> @.partitionnum);
> BEGIN TRY
> EXEC (@.command);
> END TRY
> BEGIN CATCH
> SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
> @.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
> STATISTICS_NORECOMPUTE=OFF) ';
> EXEC (@.command);
> END CATCH
> PRINT (@.command);
> END;
> PRINT 'Executed ' + @.command;
> '----
--';
> FETCH NEXT FROM partitions INTO @.objectid, @.indexid, @.partitionnum,
> @.frag;
> END;
> -- Close and deallocate the cursor.
> CLOSE partitions;
> DEALLOCATE partitions;
>
Alter Index issue & Try/Catch question
All,
I'm currently modifying a BOL procedure to rebuild/reorganize my
indexes. I've changed the ALTER INDEX command so that it performs the
rebuild online. The first time I ran the script, it errored with the
following error below:
Msg 2725, Level 16, State 2, Line 1
Online index operation cannot be performed for index
'Company$Attachment$0' because the index contains column 'Entry Pointer
ID' of data type text, ntext, image, varchar(max), nvarchar(max),
varbinary(max) or xml. For non-clustered index the column could be an
include column of the index, for clustered index it could be any column
of the table. In case of drop_existing the column could be part of new
or old index. The operation must be performed offline.
At this point, I have tried to integrate the TRY/CATCH routine so that
when this error appears, the script executes the ALTER INDEXES offline
instead (script below). I was wondering if there is a way to possible
write out the error to a log file perhaps?
Thanks,
Ian
SET NOCOUNT ON;
DECLARE @.objectid int;
DECLARE @.indexid int;
DECLARE @.partitioncount bigint;
DECLARE @.schemaname sysname;
DECLARE @.objectname sysname;
DECLARE @.indexname sysname;
DECLARE @.partitionnum bigint;
DECLARE @.partitions bigint;
DECLARE @.frag float;
DECLARE @.command varchar(8000);
-- ensure the temporary table does not exist
IF EXISTS (SELECT name FROM sys.objects WHERE name = 'work_to_do')
DROP TABLE work_to_do;
-- conditionally select from the function, converting object and index
IDs to names.
SELECT
object_id AS objectid,
index_id AS indexid,
partition_number as partitionnum,
avg_fragmentation_in_percent as frag
INTO work_to_do
FROM sys.dm_db_index_physical_stats (5, NULL, NULL , NULL, 'LIMITED')
WHERE avg_fragmentation_in_percent > 10.0 AND index_id > 0;
-- Declare the cursor for the list of partitions to be processed.
DECLARE partitions CURSOR FOR SELECT * FROM work_to_do;
-- Open the cursor.
OPEN partitions;
-- Loop through the partitions.
FETCH NEXT
FROM partitions
INTO @.objectid, @.indexid, @.partitionnum, @.frag;
WHILE @.@.FETCH_STATUS = 0
BEGIN;
SELECT @.objectname = o.name, @.schemaname = s.name
FROM sys.objects AS o
JOIN sys.schemas as s ON s.schema_id = o.schema_id
WHERE o.object_id = @.objectid;
SELECT @.indexname = name
FROM sys.indexes
WHERE object_id = @.objectid AND index_id = @.indexid;
SELECT @.partitioncount = count (*)
FROM sys.partitions
WHERE object_id = @.objectid AND index_id = @.indexid;
-- 30 is an arbitrary decision point at which to switch between
reorganizing and rebuilding
IF @.frag < 30.0
BEGIN;
SELECT @.command = 'ALTER INDEX [' + @.indexname + '] ON ' + '[' +
@.objectname + '] REORGANIZE';
IF @.partitioncount > 1
SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
@.partitionnum);
PRINT (@.command);
EXEC (@.command);
END;
IF @.frag >= 30.0
BEGIN;
SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
@.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
STATISTICS_NORECOMPUTE=OFF) ';
IF @.partitioncount > 1
SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
@.partitionnum);
BEGIN TRY
EXEC (@.command);
END TRY
BEGIN CATCH
SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
@.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
STATISTICS_NORECOMPUTE=OFF) ';
EXEC (@.command);
END CATCH
PRINT (@.command);
END;
PRINT 'Executed ' + @.command;
PRINT
'----';
FETCH NEXT FROM partitions INTO @.objectid, @.indexid, @.partitionnum,
@.frag;
END;
-- Close and deallocate the cursor.
CLOSE partitions;
DEALLOCATE partitions;To write a event log entry use: RAISERROR
RAISERROR ('Something happened and needs to be logged', 10, 1 ) WITH LOG
--
Arnie Rowland, Ph.D.
Westwood Consulting, Inc
Most good judgment comes from experience.
Most experience comes from bad judgment.
- Anonymous
You can't help someone get up a hill without getting a little closer to the
top yourself.
- H. Norman Schwarzkopf
<theredmiata@.hotmail.com> wrote in message
news:1165272932.638685.169330@.n67g2000cwd.googlegroups.com...
> All,
> I'm currently modifying a BOL procedure to rebuild/reorganize my
> indexes. I've changed the ALTER INDEX command so that it performs the
> rebuild online. The first time I ran the script, it errored with the
> following error below:
> Msg 2725, Level 16, State 2, Line 1
> Online index operation cannot be performed for index
> 'Company$Attachment$0' because the index contains column 'Entry Pointer
> ID' of data type text, ntext, image, varchar(max), nvarchar(max),
> varbinary(max) or xml. For non-clustered index the column could be an
> include column of the index, for clustered index it could be any column
> of the table. In case of drop_existing the column could be part of new
> or old index. The operation must be performed offline.
> At this point, I have tried to integrate the TRY/CATCH routine so that
> when this error appears, the script executes the ALTER INDEXES offline
> instead (script below). I was wondering if there is a way to possible
> write out the error to a log file perhaps?
> Thanks,
> Ian
>
> SET NOCOUNT ON;
> DECLARE @.objectid int;
> DECLARE @.indexid int;
> DECLARE @.partitioncount bigint;
> DECLARE @.schemaname sysname;
> DECLARE @.objectname sysname;
> DECLARE @.indexname sysname;
> DECLARE @.partitionnum bigint;
> DECLARE @.partitions bigint;
> DECLARE @.frag float;
> DECLARE @.command varchar(8000);
> -- ensure the temporary table does not exist
> IF EXISTS (SELECT name FROM sys.objects WHERE name = 'work_to_do')
> DROP TABLE work_to_do;
> -- conditionally select from the function, converting object and index
> IDs to names.
> SELECT
> object_id AS objectid,
> index_id AS indexid,
> partition_number as partitionnum,
> avg_fragmentation_in_percent as frag
> INTO work_to_do
> FROM sys.dm_db_index_physical_stats (5, NULL, NULL , NULL, 'LIMITED')
> WHERE avg_fragmentation_in_percent > 10.0 AND index_id > 0;
> -- Declare the cursor for the list of partitions to be processed.
> DECLARE partitions CURSOR FOR SELECT * FROM work_to_do;
> -- Open the cursor.
> OPEN partitions;
> -- Loop through the partitions.
> FETCH NEXT
> FROM partitions
> INTO @.objectid, @.indexid, @.partitionnum, @.frag;
> WHILE @.@.FETCH_STATUS = 0
> BEGIN;
> SELECT @.objectname = o.name, @.schemaname = s.name
> FROM sys.objects AS o
> JOIN sys.schemas as s ON s.schema_id = o.schema_id
> WHERE o.object_id = @.objectid;
> SELECT @.indexname = name
> FROM sys.indexes
> WHERE object_id = @.objectid AND index_id = @.indexid;
> SELECT @.partitioncount = count (*)
> FROM sys.partitions
> WHERE object_id = @.objectid AND index_id = @.indexid;
> -- 30 is an arbitrary decision point at which to switch between
> reorganizing and rebuilding
> IF @.frag < 30.0
> BEGIN;
> SELECT @.command = 'ALTER INDEX [' + @.indexname + '] ON ' + '[' +
> @.objectname + '] REORGANIZE';
> IF @.partitioncount > 1
> SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
> @.partitionnum);
> PRINT (@.command);
> EXEC (@.command);
> END;
> IF @.frag >= 30.0
> BEGIN;
> SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
> @.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
> STATISTICS_NORECOMPUTE=OFF) ';
> IF @.partitioncount > 1
> SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
> @.partitionnum);
> BEGIN TRY
> EXEC (@.command);
> END TRY
> BEGIN CATCH
> SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
> @.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
> STATISTICS_NORECOMPUTE=OFF) ';
> EXEC (@.command);
> END CATCH
> PRINT (@.command);
> END;
> PRINT 'Executed ' + @.command;
> PRINT
> '----';
> FETCH NEXT FROM partitions INTO @.objectid, @.indexid, @.partitionnum,
> @.frag;
> END;
> -- Close and deallocate the cursor.
> CLOSE partitions;
> DEALLOCATE partitions;
>
I'm currently modifying a BOL procedure to rebuild/reorganize my
indexes. I've changed the ALTER INDEX command so that it performs the
rebuild online. The first time I ran the script, it errored with the
following error below:
Msg 2725, Level 16, State 2, Line 1
Online index operation cannot be performed for index
'Company$Attachment$0' because the index contains column 'Entry Pointer
ID' of data type text, ntext, image, varchar(max), nvarchar(max),
varbinary(max) or xml. For non-clustered index the column could be an
include column of the index, for clustered index it could be any column
of the table. In case of drop_existing the column could be part of new
or old index. The operation must be performed offline.
At this point, I have tried to integrate the TRY/CATCH routine so that
when this error appears, the script executes the ALTER INDEXES offline
instead (script below). I was wondering if there is a way to possible
write out the error to a log file perhaps?
Thanks,
Ian
SET NOCOUNT ON;
DECLARE @.objectid int;
DECLARE @.indexid int;
DECLARE @.partitioncount bigint;
DECLARE @.schemaname sysname;
DECLARE @.objectname sysname;
DECLARE @.indexname sysname;
DECLARE @.partitionnum bigint;
DECLARE @.partitions bigint;
DECLARE @.frag float;
DECLARE @.command varchar(8000);
-- ensure the temporary table does not exist
IF EXISTS (SELECT name FROM sys.objects WHERE name = 'work_to_do')
DROP TABLE work_to_do;
-- conditionally select from the function, converting object and index
IDs to names.
SELECT
object_id AS objectid,
index_id AS indexid,
partition_number as partitionnum,
avg_fragmentation_in_percent as frag
INTO work_to_do
FROM sys.dm_db_index_physical_stats (5, NULL, NULL , NULL, 'LIMITED')
WHERE avg_fragmentation_in_percent > 10.0 AND index_id > 0;
-- Declare the cursor for the list of partitions to be processed.
DECLARE partitions CURSOR FOR SELECT * FROM work_to_do;
-- Open the cursor.
OPEN partitions;
-- Loop through the partitions.
FETCH NEXT
FROM partitions
INTO @.objectid, @.indexid, @.partitionnum, @.frag;
WHILE @.@.FETCH_STATUS = 0
BEGIN;
SELECT @.objectname = o.name, @.schemaname = s.name
FROM sys.objects AS o
JOIN sys.schemas as s ON s.schema_id = o.schema_id
WHERE o.object_id = @.objectid;
SELECT @.indexname = name
FROM sys.indexes
WHERE object_id = @.objectid AND index_id = @.indexid;
SELECT @.partitioncount = count (*)
FROM sys.partitions
WHERE object_id = @.objectid AND index_id = @.indexid;
-- 30 is an arbitrary decision point at which to switch between
reorganizing and rebuilding
IF @.frag < 30.0
BEGIN;
SELECT @.command = 'ALTER INDEX [' + @.indexname + '] ON ' + '[' +
@.objectname + '] REORGANIZE';
IF @.partitioncount > 1
SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
@.partitionnum);
PRINT (@.command);
EXEC (@.command);
END;
IF @.frag >= 30.0
BEGIN;
SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
@.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
STATISTICS_NORECOMPUTE=OFF) ';
IF @.partitioncount > 1
SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
@.partitionnum);
BEGIN TRY
EXEC (@.command);
END TRY
BEGIN CATCH
SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
@.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
STATISTICS_NORECOMPUTE=OFF) ';
EXEC (@.command);
END CATCH
PRINT (@.command);
END;
PRINT 'Executed ' + @.command;
'----';
FETCH NEXT FROM partitions INTO @.objectid, @.indexid, @.partitionnum,
@.frag;
END;
-- Close and deallocate the cursor.
CLOSE partitions;
DEALLOCATE partitions;To write a event log entry use: RAISERROR
RAISERROR ('Something happened and needs to be logged', 10, 1 ) WITH LOG
--
Arnie Rowland, Ph.D.
Westwood Consulting, Inc
Most good judgment comes from experience.
Most experience comes from bad judgment.
- Anonymous
You can't help someone get up a hill without getting a little closer to the
top yourself.
- H. Norman Schwarzkopf
<theredmiata@.hotmail.com> wrote in message
news:1165272932.638685.169330@.n67g2000cwd.googlegroups.com...
> All,
> I'm currently modifying a BOL procedure to rebuild/reorganize my
> indexes. I've changed the ALTER INDEX command so that it performs the
> rebuild online. The first time I ran the script, it errored with the
> following error below:
> Msg 2725, Level 16, State 2, Line 1
> Online index operation cannot be performed for index
> 'Company$Attachment$0' because the index contains column 'Entry Pointer
> ID' of data type text, ntext, image, varchar(max), nvarchar(max),
> varbinary(max) or xml. For non-clustered index the column could be an
> include column of the index, for clustered index it could be any column
> of the table. In case of drop_existing the column could be part of new
> or old index. The operation must be performed offline.
> At this point, I have tried to integrate the TRY/CATCH routine so that
> when this error appears, the script executes the ALTER INDEXES offline
> instead (script below). I was wondering if there is a way to possible
> write out the error to a log file perhaps?
> Thanks,
> Ian
>
> SET NOCOUNT ON;
> DECLARE @.objectid int;
> DECLARE @.indexid int;
> DECLARE @.partitioncount bigint;
> DECLARE @.schemaname sysname;
> DECLARE @.objectname sysname;
> DECLARE @.indexname sysname;
> DECLARE @.partitionnum bigint;
> DECLARE @.partitions bigint;
> DECLARE @.frag float;
> DECLARE @.command varchar(8000);
> -- ensure the temporary table does not exist
> IF EXISTS (SELECT name FROM sys.objects WHERE name = 'work_to_do')
> DROP TABLE work_to_do;
> -- conditionally select from the function, converting object and index
> IDs to names.
> SELECT
> object_id AS objectid,
> index_id AS indexid,
> partition_number as partitionnum,
> avg_fragmentation_in_percent as frag
> INTO work_to_do
> FROM sys.dm_db_index_physical_stats (5, NULL, NULL , NULL, 'LIMITED')
> WHERE avg_fragmentation_in_percent > 10.0 AND index_id > 0;
> -- Declare the cursor for the list of partitions to be processed.
> DECLARE partitions CURSOR FOR SELECT * FROM work_to_do;
> -- Open the cursor.
> OPEN partitions;
> -- Loop through the partitions.
> FETCH NEXT
> FROM partitions
> INTO @.objectid, @.indexid, @.partitionnum, @.frag;
> WHILE @.@.FETCH_STATUS = 0
> BEGIN;
> SELECT @.objectname = o.name, @.schemaname = s.name
> FROM sys.objects AS o
> JOIN sys.schemas as s ON s.schema_id = o.schema_id
> WHERE o.object_id = @.objectid;
> SELECT @.indexname = name
> FROM sys.indexes
> WHERE object_id = @.objectid AND index_id = @.indexid;
> SELECT @.partitioncount = count (*)
> FROM sys.partitions
> WHERE object_id = @.objectid AND index_id = @.indexid;
> -- 30 is an arbitrary decision point at which to switch between
> reorganizing and rebuilding
> IF @.frag < 30.0
> BEGIN;
> SELECT @.command = 'ALTER INDEX [' + @.indexname + '] ON ' + '[' +
> @.objectname + '] REORGANIZE';
> IF @.partitioncount > 1
> SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
> @.partitionnum);
> PRINT (@.command);
> EXEC (@.command);
> END;
> IF @.frag >= 30.0
> BEGIN;
> SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
> @.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
> STATISTICS_NORECOMPUTE=OFF) ';
> IF @.partitioncount > 1
> SELECT @.command = @.command + ' PARTITION=' + CONVERT (CHAR,
> @.partitionnum);
> BEGIN TRY
> EXEC (@.command);
> END TRY
> BEGIN CATCH
> SELECT @.command = 'ALTER INDEX [' + @.indexname +'] ON ' + '[' +
> @.objectname + '] REBUILD WITH (ONLINE=OFF, SORT_IN_TEMPDB=ON,
> STATISTICS_NORECOMPUTE=OFF) ';
> EXEC (@.command);
> END CATCH
> PRINT (@.command);
> END;
> PRINT 'Executed ' + @.command;
> '----';
> FETCH NEXT FROM partitions INTO @.objectid, @.indexid, @.partitionnum,
> @.frag;
> END;
> -- Close and deallocate the cursor.
> CLOSE partitions;
> DEALLOCATE partitions;
>
Wednesday, March 7, 2012
Alter column question
I've been asked to consider modifying our largest, oldest and most used database. "They" want to change the type of some columns from tinyint & smallint to dec(4,1) thru dec(6,1).
I ran one of these commands on a local subset and it took forever to finish. The full script will have to be done over the weekend.
I'm doing all I can to prevent this asinine change, mentioning the fact that everything we've built in the last 12 years will need to be checked / modified.
But, in case I lose, I wrote a script to modify the columns. It consists of a bunch of T-SQL commands like:
alter table course_table
alter column credit_hours decimal(4,1)
go
I ran one of these commands on a local subset and it took forever to finish. The full script will have to be done over the weekend.
So the question is, what is happening to the logs when this command is executing? Should I try dumping it after each alter column?
Any other advice on this subject to offer?
Thanks.
Depending on the type of change, ALTER TABLE will result in just metadata updates or it has to rewrite every single row. In your case, it will rewrite every single row and the time it takes is directly proportional to the number of rows/row size/data pages. The ALTER TABLE itself is atomic in nature so you can't do much in terms of reducing the logging resources for it. So it will log every change in your case to the log. But you can take a log backup after each ALTER TABLE or periodically to manage the log growth. See the link below for some details on the ALTER TABLE also:
http://www.sqlmag.com/Article/ArticleID/40538/Inside_ALTER_TABLE.html
|||That's what I was suspecting about the logs.Thanks for the quick reply.
Subscribe to:
Posts (Atom)