Jdbc and paging

Question

Hi,
i have the following doubts regarding the jdbc and paging please clarify me...
1)Is it advisable to use stored procedure to implement paging is records
are more in database
2)If the records are more in the database and we execute
select * from table where id >10 and id<15 then will
the query execution will be slow because the database is large(and so stored procedure will be faster)
or as we are selecting only records between 10 and 15 so the
query execution is not slow
3)will the recordset help in solving the probem
4)is there a way to select the count of the records along with the records for display in the page
the count of the records is required to display the next button
if two queries are to be written one to display the next button and second to get the records to
display on the page then the system may be slow when many users are accessing and using stored
procedures may be better
5)when we write the select query with 'for update' clause then the row will be locked
or the table is locked .
6)when we write the select query with 'for update' clause then until we execute the update
query the lock will be there and the lock will be released on executing the update query
is that right then what about the case if after the select query we get some exception
and code does not reach the update query then what about the lock will be lock remain
and give problem
regards
Javid

Answer 1

1) that would really depend on you goals. for performance only, stored procedures are precompiled so should be faster.
2) definitely you should only select what you want and limiting the number of rows you return will speed up your query.
3) what do you mean by that? a recordset represents the collection of rows returned from the query so should help you.
4) it depends on your database. for instance, MySQL has a function to return the number of rows returned. however, some database has an unbuffered mode so that the first few rows will be returned very quickly without knowing how many rows are there, in which case, you won't know the number of rows until all the rows have been retrieved.
5 and 6 are database-dependent, while most rdbms only lock the selected rows by default, some lock the entire table.
Java 2 MVP
http://www.brainbench.com