Hi!
On Nov 29, 2008, at 6:14 AM, Roland Bouman wrote:
Not rigurously. But from what I recall, for a single thread (which is
the case I am interested in) inserting something like 10 mln rows was
fastest on CSV and somewhat (5-10%) slower with load data infile.
It is surprising that it is that much quicker. The difference in code
between the two is not a lot. I never did a benchmark when we went
from using mmap() to read() with CSV... so it might have been faster
at one point for that reason.
Cheers,
-Brian
--
_______________________________________________________
Brian "Krow" Aker, brian at tangent.org
Seattle, Washington
http://krow.net/ <-- Me
http://tangent.org/ <-- Software
_______________________________________________________
You can't grep a dead tree.
_______________________________________________
Mailing list: https://launchpad.net/~drizzle-discuss
Post to : [email protected]
Unsubscribe : https://launchpad.net/~drizzle-discuss
More help : https://help.launchpad.net/ListHelp
