when volume of data changes so change plans - sometimes its just that.

could you supply us with the new explain plan - maybe that will tell us why but fact...

lets see if I can

a recursive CTE works by having a "anchor" set - that is the sql before the union. - this is the start point for all records...

I think that you can do that using another CTE.

and I do advise the use of () around the formula blocks so there is no ambiguity of what is calculated...

google for unpivot - or use a outer apply selecting each column and its value with a case statement and a group by

when you have some code showing what you...

• This reply was modified 5 years, 6 months ago by frederico_fonseca.

SSDT is for the client machine to develop the packages - Not to be installed on the server

Note that to load CSV/Excel files you do not need SSDT - that...

if you are working with SQL 2019 and get R configured and up and running you can write Excel files with it - and probably easier than with SSIS

There was...

• This reply was modified 5 years, 6 months ago by frederico_fonseca.
• This reply was modified 5 years, 6 months ago by frederico_fonseca.

did you try cleaning the query and using exists instead of not in ?

Assuming that the query is exactly as you supplied the following could be changed - may or...

Adding to the sql from Chris above.

you wanna search for any module that has

• bulk insert
• openrowset
• openquery
• xp_cmdshell
• oa_*** extended sp's ( OLE Automation)

you also need to look for...

if only the where is generated but remaining query remains the same maybe changing it slightly could influence QE to choose a different path.

note that for the following the index...

Something that seems a bit obvious to me but hasn't actually been posted in this thread is that you really need to simplify your proc. Anything that is 3300...

link is to allow you to give permissions to start a job - it has nothing to do with what the job does afterwards.

so separating the 2 things - did...

only changes to "character size" was done on 2012 with the inclusion of new _SC collations to allow for UTF-16 and the expansion done by 2019 to allow for char...

not needed any more

• This reply was modified 5 years, 6 months ago by frederico_fonseca.

you are missing the commas separating the columns

thanks for the clarification.

this should then do the trick - as before replace the "base" table with your own table

select b1.EMID
     , b1.visit_date_01
...