My problem is the following:
In my vb5 project, I have two longinteger arrays of dimension n (n between
10 to 1000), which I would like to store in a table in SQL-Server 6.5 via
If I run the INSERT query n times passing the contents of the arrays row by
row, it takes about 15 seconds for n=1000. If I generate a stored procedure
which takes 2*m parameters, so that I can store m rows in a single query, it
takes exactly the same time to store the 1000 rows in the table.
An idea I had was to pass all data in the arrays as varbinary parameter and
then to extract the information with substring in the stored procedure, so I
would have only one query to call and would call it assynchronously. But
this would not work because I did not succeed in storing a whole array in
one rdoQuery parameter from vb. Another idea would consist in passing the
information in a variable of type text. This is not straight forward because
I could not manage to extract parts of the variable to use as value for the
Now my questions:
1) Can I expect performance gains by limiting the number of queries for my
1000 INSERT statements or is the bottleneck in the INSERT itself?
2) Is there a way of storing the contents of an array of longint or an array
of byte in an rdoQuery parameter of type varbinary?
3) How could I possibly extract parts of the contents of variable of type
4) Has anybody some better idea to solve my problem?
Thanks a lot for your help.