This is a 7.2 database. I am trying to test this, but having resource
troubles, and running out of time. I want to delete rows from
myschema.mytable1, but only the rows that meet the following condition. I
have structured the SQL using a correlated reference. Is this right, or
will I delete more than meets the subquery condition ?


delete from a where
(select a.host_id, a.field2, a.field3, a.field4 from
myschema.mytable1 a,
myschema.mytable2 b
where
a.host_id=b.host_id and a.field2=b.field2 and
b.field3=b.field3 and a.field4=b.field4)


Thanks,
Scott

Hi Scott,

The correct syntax is:

delete
from myschema.mytable1 a
where exists
(select *
from myschema.mytable2 b
where a.host_id=b.host_id
and a.field2=b.field2
and b.field3=b.field3
and a.field4=b.field4)

If you replace the "delete" with a "select *", you can see exactly
which rows will be deleted before going ahead with the delete. If you
are expecting a large number of rows Select count(*) will tell you how
many rows will be deleted.

Christian.