I would store this kind of ranking in separate column and update this column using trigger when data in the table is changed.

This update not suppose to happen to often,...

sol,

how many records WITHOUT distinct (see Jeff's query) and how "heavy" is each of those records?

I mean what is aggregate size of all columns in your recordset?

Locks really depend on data statistics, indexing, etc.

Someone dropped or changed an index on this table, statistics forced Server to choose another execution plan...

It's not so easy to guess,...

Because when you have INSERT outside the whole SP appears locked in transaction (INSERT is implicit transaction, right?)

It does not happen if you perform INSERT inside of SP.

Try this workaround:

SELECT * INTO #agingData

FROM dbo.agingData

WHERE 1=0 -- copies table structure, no data copied

INSERT #agingData

EXEC dbo.populateAging

INSERT INTO dbo.agingData

SELECT * from #agingData

If it works it's definitely...

Jason,

do you read table dbo.agingData inside SP?

It appears locking problem.

If you perform INSERT ... EXEC you lock targeted table. If you are trying to read something from this table...

CREATE TRIGGER Trig_vendor_code ON [vendor] FOR UPDATE

AS

UPDATE V

SET vendorcode = i.vendorcode

FROM dbo.Vendor V

INNER JOIN inserted i ON V.origvenid = i.vendorID

WHERE V.vendorcode i.vendorcode -- add NULL checks if...

CREATE PROC dbo.Blah-blah ..., @NAME varchar(255), ....

AS

...

SET @NAME = replace ( @NAME,char(39),char(39)+char(39))

...

GO

DISTINCT is simplest option of GROUP BY.

It makes SQL Server rescan returning resultset and group the data.

It involves creating hash table in tempbd for holding temporary data.

Your query must return...

Breaking relational data model and normakization rules is not smart.

It does not simplify your life.

Homogenious data must be in a single column. Period.

And column names [1]...[18] must become numbers in...

Look for DATEPART in BOL.

Check cross links there to be aware of other datetime functions.

I don't see any problems with EAV for those who does not use to use SELECT *

In poperly designed application you need to return FirstName, LastName, etc. once per person....

Join to CALENDAR table.

I know the way.

Normalize your data and get rid of fat tables.

One view could solve all your problems with joins.