a way to insert from a range of rows

[jayvaughn]

I'm needing to do an insert (selecting rows from another table), but only want to specify a certain range of rows.

If I do this...

select * from
(select dsln, row_number() over() as rownumber
from ldatacpd.srvdtrn) as column
where rownumber between 1 and 3

then i am able to select the rows from the source table as desired.

However if I want to morph that into an insert statement such as the following, i run into issues.

insert into ls#jhv.srvdtrn (dsln)
(select * from
(select dsln, row_number() over() as rownumber
from ldatacpd.srvdtrn) as column
where rownumber between 1 and 3)

Any suggestions on how to restructure and overcome?

Guess I could always build a cursor from the row numbers selected then do individual inserts for each row, but I'm looking for the best performance option.

[Vectorspace]

It is because your outer select query is returning two fields - dsln and rownumber, but the insert is expecting only one (dsln)

Try this (difference in bold)
insert into ls#jhv.srvdtrn (dsln)
(select dsln from
(select dsln, row_number() over() as rownumber
from ldatacpd.srvdtrn) as column
where rownumber between 1 and 3)

If you only want the first three rows (as you have coded) as opposed to any random range of rows, then you could also do this:
insert into ls#jhv.srvdtrn (dsln)
(select dsln from ldatacpd.srvdtrn
fetch first 3 rows only) as column