[Ruby-DBI-devel] is select_all lazy?

[Steven R. <jun...@sh...>]

Hi,

I'm using DBI with MySQL, and am having a problem using the block version 
of select_all with a large table.

If I do
    dbh.select_all('select * from small_table') do |row|
      p row
    end
it works fine.

If I do
    dbh.select_all('select * from big_table limit 10') do |row|
      p row
    end
it works fine.

If I do
    dbh.select_all('select * from big_table') do |row|
      p row
    end
(where big_table is 25 gigs) it hangs without executing the block, and
the mysqld server process pegs the CPU.

I'm guessing that the block version of select_all is fetching the entire 
result set instead of fetching rows lazily (i.e., it's equivalent to 
dbh.select_all(...).each do |row|). Please tell me this isn't the case...

thanks a lot,
steven.