should i expect data types in task->export data from ssms?

• stan

  Ten Centuries

  Points: 1140

  More actions

  June 27, 2025 at 3:12 pm

  #4614248

  Hi i recently exported a number of "to be purged resultsets " to .tsv's.     this is a related link  https://www.sqlservercentral.com/forums/topic/is-there-a-reorg-available-in-sql-server .

  my next step was to see how i'd fare trying to rehydrate to a table with the same cols and data types from one of these tsv's, specifically one with high volume (9 million records at 500+ bytes per record) .

  i was surprised by a few things and may be disappointed now that i dint listed to the sme's who offered opinions in that other link...

  1.  what was a varchar (10) column way at the source looked like a varchar(50) to the import (tasks->import...) and the process aborted
  2.  at least one of the floats from way back at the source looked like a string to the import and therefore generated a conversion error in the import
  3.  choosing instead "import from a flat file" which presumably means I (sql server) will create the destination table for you went into one of those "program is not responding" states.
  4. if you click outside the "focus" pop up window where "progress"  is showing, but still on the rdp window (if you rdp'd) , the process stops.   Thats kind of disturbing because had i not compared record counts i never would have known.

  this is on a vm with limited memory and other resources.   does the community believe that if i am to even have a chance i'll either have to target the output table  from option 1,2 as one with all varchar data types or get in there and edit each mapping?   what is that global conversion i see where you can choose ignore or abort instead?

  • This topic was modified 3 weeks, 1 day ago by stan.
  • This topic was modified 2 weeks, 5 days ago by stan.

• stan

  Ten Centuries

  Points: 1140

  More actions

  June 27, 2025 at 5:46 pm

  #4614313

  possible correction, testing now, the target fields didnt exactly match up , maybe the import expects them to match in spite of the column headings.   i'll post back here.

  Indeed,  for some reason "floats" dont work in reverse if the goal is to map them  back from the .tsv's to a schema with data types of float.   This is kind of disturbing because if i'm not mistaken i've seen sql able to do that in an ssis pkg, but to be honest im not 100% sure.

  • This reply was modified 2 weeks, 5 days ago by stan.