option 1 is incorrect - just discard it.

can you give the plan for the second? was expecting it to be a bit faster than that.

regarding your last query (from the second plan). two options below

option 1 is just a adjustment of your original query - does same thing but with 1 less join. may...

can you try this slight variation of my second example and give runtime and plan as well.

drop table if exists #weeklastday;
drop table if exists #temp1;

select distinct t1.EXPIRATION
into...

that plan is from you version of the change e.g. where you still keep the dateadd and your temp table does not filter or change the expiration date.

can you give...

how much faster? can can you also put new explain plan.

small error on my original code - the temp table was referring an non existing alias (t1)- easy to fix I think.

another attempt at simplifying it assuming that I correctly...

no difference.

your code was only looking for the the friday date on a particular situation e.g. when the EXPIRATION = Thursday

so by determining all the FRIDAY dates in advance, subtracting...

adding further indexes on this case likely not the best - the underlying table is a columnstore table.

the thing that would speed up the most is to have a computed...

a image of the plans don't help much - please post the real explain plan - actual if possible.

use https://www.brentozar.com/pastetheplan/ for it

SSIS has behind the scenes to retrieve the table metadata - this error seems to indicate a mismatch between the server and the client driver - or some parameters need...

this is one of those cases where you do NOT want to use SSIS native dataflow to do the transfer.

<li class="text--left" style="text-align: left;">SSIS Script - using .net sql datareader...

I would not consider it a big issue for now - you can still install prior versions of SSMS and keep using it.

does it work if you try

DECLARE @ids NVARCHAR(max)
select @ids = string_agg(inserted.[lo_location], ',')
from (select X.query('.').value('(row/lo_Location)[1]', 'int') as lo_location
          ...

change your select to include all columns from the commented out join - and uncomment it out as well and order by variant code

then look at the results and see...

what is the aversion at googling before posting? a simple google for "powershell wait" returns you the command you need to use for it.