estimate size of table based on number of rows

Hi everyone,
This might be a tough one, I don't know if this is doable.
Basically I want to create a stored procedure to return estimated size of a
table.
because I don't know how many rows it will have in the future, I want to
calcuate how much disk space it takes to have one row,
then multiply by number of rows I specified. (space calculation for index is
not necessary).
EXEC GetEstimateTableSize @.TableName='Table1', @.NumberOfRows ='3000000'
it'll return value in KBs after I execute proc. possible?If you run :
EXEC dbo.sp_spaceused table_name
You get the current space usage. Divide it by the current number of rows,
and multiply by the projected one.
If the table is completely empty or you'd rather calculate theoretical size,
there are formulas you can use from BOL, or better yet, a lengthy discussion
on internal structures and row sizes in Inside SQL Server 2000.
BG, SQL Server MVP
www.SolidQualityLearning.com
"Britney" <britneychen_2001@.yahoo.com> wrote in message
news:uWKIMMBpFHA.3936@.TK2MSFTNGP10.phx.gbl...
> Hi everyone,
> This might be a tough one, I don't know if this is doable.
> Basically I want to create a stored procedure to return estimated size of
> a table.
> because I don't know how many rows it will have in the future, I want to
> calcuate how much disk space it takes to have one row,
> then multiply by number of rows I specified. (space calculation for index
> is not necessary).
>
> EXEC GetEstimateTableSize @.TableName='Table1', @.NumberOfRows ='3000000'
>
> it'll return value in KBs after I execute proc. possible?
>
>
>
>
>
>|||Have you looked in the Books Online for sp_spaceused? It'll get you
part of the way there because it returns the rows used and the current
size of the table. Here's a quick and dirty stab at it; obviously
it'll need polishing:
This is actually a pretty useful idea; I'm planning on using this
myself.
Stu
DECLARE @.Table varchar(255)
DECLARE @.NumberOfRows int
SET @.Table = 'Splat'
SET @.NumberOfRows = 1
CREATE TABLE #t (name varchar(255),
rows int,
reserved varchar(100),
data varchar(100),
index_size varchar(100),
unused varchar(100))
--how big is the table now?
exec sp_spaceused @.Table
INSERT INTO #t
exec sp_spaceused @.Table
--strip off the ' KB' from the data column
--convert data and rows to decimal, and
--divide data by number of rows and multiply by number of anticipated
rows
SELECT rows, data, (data/rows) * @.NumberOfRows
FROM ( SELECT rows = CONVERT(decimal(32,3), rows),
data = CONVERT(decimal(32,3), LEFT(data, LEN(data)-3))
FROM #t) x
DROP TABLE #t
HTH,
Stu|||The following may help:
http://www.microsoft.com/downloads/...&displaylang=en
If you need it in a SP, try to understand the formulas used in the
spreadsheet and translate them in T-SQL (using the data from syscolumns
and other system tables).
Razvan|||I see 4 columns called reserved , index_size, unused, data
I guess I need to add 4 columns to get the total size, then divide by number
of row to find out how much disk space per row?
then disk space per row * estimate number of rows to find out estimate size?
"Itzik Ben-Gan" <itzik@.REMOVETHIS.SolidQualityLearning.com> wrote in message
news:etjHceBpFHA.2888@.TK2MSFTNGP10.phx.gbl...
> If you run :
> EXEC dbo.sp_spaceused table_name
> You get the current space usage. Divide it by the current number of rows,
> and multiply by the projected one.
> If the table is completely empty or you'd rather calculate theoretical
> size, there are formulas you can use from BOL, or better yet, a lengthy
> discussion on internal structures and row sizes in Inside SQL Server 2000.
> --
> BG, SQL Server MVP
> www.SolidQualityLearning.com
>
> "Britney" <britneychen_2001@.yahoo.com> wrote in message
> news:uWKIMMBpFHA.3936@.TK2MSFTNGP10.phx.gbl...
>|||You need the data + index_size.
Reserved includes: data + index_size + unused.
BG, SQL Server MVP
www.SolidQualityLearning.com
"Britney" wrote:

> I see 4 columns called reserved , index_size, unused, data
> I guess I need to add 4 columns to get the total size, then divide by numb
er
> of row to find out how much disk space per row?
> then disk space per row * estimate number of rows to find out estimate siz
e?
> "Itzik Ben-Gan" <itzik@.REMOVETHIS.SolidQualityLearning.com> wrote in messa
ge
> news:etjHceBpFHA.2888@.TK2MSFTNGP10.phx.gbl...
>
>|||but I have a NTEXT column,
I don't think it can calculate NTEXT.
"Razvan Socol" <rsocol@.gmail.com> wrote in message
news:1124386169.690259.289640@.z14g2000cwz.googlegroups.com...
> The following may help:
> http://www.microsoft.com/downloads/...&displaylang=en
> If you need it in a SP, try to understand the formulas used in the
> spreadsheet and translate them in T-SQL (using the data from syscolumns
> and other system tables).
> Razvan
>|||Indeed, ntext columns are not covered by the spreadsheed.
In the DataSizer.doc file, they wrote:
The tool does not include the formula to estimate the size of a table
that has Text columns. Not NULL text values consume 16 bytes in the
data row and have a minimum size of 84 bytes on the text page. Text
values are packed onto text pages with the same algorithm as data
rows so it should be possible to estimate the size of text data
storage using the HEAP table spreadsheet if you know the average size
of your text values.
Razvan|||if I use sp_spaceused 'tablename' command,
if that table has a few NText Columns,
I think it will calculate total spaces including disk spaces for NTEXT
column, correct?
"Razvan Socol" <rsocol@.gmail.com> wrote in message
news:1124483366.622335.202490@.g49g2000cwa.googlegroups.com...
> Indeed, ntext columns are not covered by the spreadsheed.
> In the DataSizer.doc file, they wrote:
> The tool does not include the formula to estimate the size of a table
> that has Text columns. Not NULL text values consume 16 bytes in the
> data row and have a minimum size of 84 bytes on the text page. Text
> values are packed onto text pages with the same algorithm as data
> rows so it should be possible to estimate the size of text data
> storage using the HEAP table spreadsheet if you know the average size
> of your text values.
> Razvan
>

ERROR_STATE() Always = 0?

Below is the contents of a SPROC I have. I want to return the error
info in the catch block for it so I call RaiseError. But the
ERROR_STATE() always comes up as 0 which is ilegal since it must be
between 1-127 I guess. So I wrote a stupid if block to set it to 1 to
stop that error.

But I'd really rather not do this. Can anyone tell what's wrong? Why am
I not getting a correct return from ERROR_STATE()?

Thanx much:

BEGIN TRY
BEGIN TRANSACTION
DELETE FROM WebUser2Role WHERE WebUserID = @.WebUserID

INSERT INTO WebUser2Role
SELECT value, @.WebUserID FROM fIntList2Table(@.RoleIDList)
END TRY
BEGIN CATCH
DECLARE @.ErrorMessage NVARCHAR(4000)
DECLARE @.ErrorSeverity INT
DECLARE @.ErrorState INT

SELECT
@.ErrorMessage = ERROR_MESSAGE(),
@.ErrorSeverity = ERROR_SEVERITY(),
@.ErrorState = ERROR_STATE();

IF @.ErrorState = 0
BEGIN
SET @.ErrorState = 1
END

IF @.@.TRANCOUNT > 0
BEGIN
ROLLBACK TRANSACTION
END

RAISERROR (@.ErrorMessage, @.ErrorSeverity, @.ErrorState )

END CATCH

IF @.@.TRANCOUNT > 0
BEGIN
COMMIT TRANSACTION
ENDThe only thing I can think of which may help is that
Errors with a severity of 10 or lower are considered warnings or
informational messages, and are not handled by TRY.CATCH blocks.
therefore maybe it's not recognised as an error ?
--
Jack Vamvas
___________________________________
Receive free SQL tips - www.ciquery.com/sqlserver.htm

<wackyphill@.yahoo.com> wrote in message
news:1143771643.520517.319500@.e56g2000cwe.googlegr oups.com...
> Below is the contents of a SPROC I have. I want to return the error
> info in the catch block for it so I call RaiseError. But the
> ERROR_STATE() always comes up as 0 which is ilegal since it must be
> between 1-127 I guess. So I wrote a stupid if block to set it to 1 to
> stop that error.
> But I'd really rather not do this. Can anyone tell what's wrong? Why am
> I not getting a correct return from ERROR_STATE()?
> Thanx much:
>
>
> BEGIN TRY
> BEGIN TRANSACTION
> DELETE FROM WebUser2Role WHERE WebUserID = @.WebUserID
> INSERT INTO WebUser2Role
> SELECT value, @.WebUserID FROM fIntList2Table(@.RoleIDList)
> END TRY
> BEGIN CATCH
> DECLARE @.ErrorMessage NVARCHAR(4000)
> DECLARE @.ErrorSeverity INT
> DECLARE @.ErrorState INT
> SELECT
> @.ErrorMessage = ERROR_MESSAGE(),
> @.ErrorSeverity = ERROR_SEVERITY(),
> @.ErrorState = ERROR_STATE();
> IF @.ErrorState = 0
> BEGIN
> SET @.ErrorState = 1
> END
> IF @.@.TRANCOUNT > 0
> BEGIN
> ROLLBACK TRANSACTION
> END
> RAISERROR (@.ErrorMessage, @.ErrorSeverity, @.ErrorState )
> END CATCH
> IF @.@.TRANCOUNT > 0
> BEGIN
> COMMIT TRANSACTION
> END|||No it deffinately is going into the catch block.

But ERROR_STATE() in the catch block returns 0. This poses a problem
when I call RaiseError because 0 is not a valid state.

error_number() won't return error code

hi all,

i have sql statement in ExecuteSQLTask, connecting to AdventureWorks

BEGIN TRY
insert into person.contacts(contactid) values (1);
END TRY
BEGIN CATCH
SELECT ERROR_NUMBER() AS ErrorNumber;
END CATCH;

i purposely put a table that does not exist in the database to catch
the error. however, the error number that is 208 won't appear, instead
the whole error message will be displayed.

if i change the query in try block to
insert into person.contact(contactid) values (1);

it'll return the error number 544, which is correct.

what did i do wrong?
thanks!

You did nothing wrong.

Try/Catch doesn't catch all errors, specifically compile time errors (there are a few others). Try running the following, "more obvious" compile time error. This compile time error won't be caught either, since the execution of the statements never happens.

BEGIN TRY

PRINT 'Inside Try-Block'

'This is not sql'

END TRY

BEGIN CATCH

SELECT error_number() as 'Error In Try Block'

END CATCH

|||

thanks jaegd for your reply.. it didn't say in BOL that try/catch doesn't catch all errors or maybe i left that out.

anyhow, my problem is not solved. is there an alternative where i can catch any errors that occur?

|||See the following BOL reference for the constructs which Try/Catch does not trap.
http://msdn2.microsoft.com/en-us/library/ms175976.aspx

However, To catch almost every error (including compile time and deferred name resolution errors), you can use dynamic SQL in combination with sp_executesql.

BEGIN TRY
DECLARE @.not_sql NVARCHAR(100)
SET @.not_sql = 'This is not sql'
EXEC sp_executesql @.not_sql

END TRY
BEGIN CATCH
SELECT ERROR_NUMBER() AS ErrorNumber
END CATCH

Note that almost every error is caught with this method, because broken client connections are not

trapped, hopefully for relatively obvious reasons (you have no

connection).|||thanks a lot jaegd