you said "  DB is set to read committed" - does this mean your db has Read Committed Snapshot Isolation on? if so this is likely why.

you have already been given some answers on your original post - https://www.sqlservercentral.com/forums/topic/eliminating-csv-conversion-step

with regards to the errors, they are standard SSIS errors, which you can both read manuals and google...

so you are asking us to do your work for free?

one thing is to help with particular issues or problems. Error you may get with code for example.

But this is...

easy - would help if you had created a script task (as instructed on the link) instead of a script component.

so do it correctly - script component does NOT allow...

in addition to what Phil supplied - which may mislead you into thinking you can set it up to bypass this particular issue.

SSIS catalog  DOES NOT allow for double hop...

and as I said on my prior post - it will not work like that as it is a double hop situation - only way for this execution to work...

the way it is done, on many companies, is to restore the DB on a PROD server (can be same server as the source if server has the required capacity)),...

let me put it in a different way.

do those packages access any server other than the one where SSIS package is installed and running, and is the connection to those...

that message is a typical windows authentication double hop issue (which SSIS Catalog does NOT support).

has NOTHING to do with permissions just to be clear.

so... when you execute the package...

this is why you should supply consumable data and desired output.

so on above example, how do we link between tbl1 and tbl2? this is assuming that there are more than...

error is very clear - your bulk insert is failing because of data conversion between input (your file) and the output (#TEMP_TABLE).

it has Nothing to do with that flag or...

its a matter of choice - but I personally do not use project connections, only project parameters.

Connections are all defined within each package that needs them - and the connection...

as Phil said - don't save sensitive parameters. e.g. set packages to "DontSaveSensitive"

while this has a downside in that sensitive parameters are not saved, there are ways around it so...

why don't you try it on dev, on a copy of the table with all required indexes? asking us here for something you can try yourself is a waste of...

• This reply was modified 1 years, 12 months ago by frederico_fonseca.

as ratbak said you will need to change the clustered index to have the partitioned column as its first (not sure if unique) column.

and in order to allow for the...