The skew buckets are dynamically
flushed just like buckets in a dynamic hash join would be.
- Bryce Cutt
On Fri, Mar 20, 2009 at 5:51 PM, Robert Haas wrote:
> On Fri, Mar 20, 2009 at 8:45 PM, Bryce Cutt wrote:
>> On Fri, Mar 20, 2009 at 5:35 PM, Robert Haas wrote:
>>> If the
way through if too many inner tuples fall into the
new "skew buckets" (formerly IM buckets) and dump the tuples back into
the main buckets. The potential win is still pretty high though.
- Bryce Cutt
On Fri, Mar 20, 2009 at 5:35 PM, Robert Haas wrote:
> On Fri, Mar 20, 2009 at 8:14
nodeHashjoin.c (in ExecHashJoin) has
been rearranged so that in the single batch case short circuit
evaluation requires only the first test in the IF to be checked.
The "limited skew" check mentioned in Case 2 above is a simple check
in the ExecHashJoinDetectSkew function.
- Bryce Cutt
en at all and instead of the if
statement happening per tuple it is run just once per join. We have
to test this a bit more but it should further reduce the overhead.
Hopefully we will have the new patch ready to go this weekend.
- Bryce Cutt
On Thu, Feb 26, 2009 at 7:45 AM, Tom Lane
on finalizing this patch, especially in
regard to style issues. Thank you for all the help.
- Dr. Ramon Lawrence and Bryce Cutt
On Sun, Jan 4, 2009 at 6:48 PM, Robert Haas wrote:
>> 1) Isn't ExecHashFreezeNextMCVPartition actually a most common TUPLE
>> partition, rather t
the probe
side is an operator other than a seq scan (such as another hashjoin)
the code can now find the stats tuple for the underlying relation.
The new idea of limiting the number of MCVs to a percentage of memory
has not been added yet.
- Bryce Cutt
On Mon, Dec 29, 2008 at 8:55 PM, Robert
ccepted.
It relies on getting the stats tuple for the join during the planning
phase (in the cost function) and estimating the benefit that would
have on the join cost.
- Bryce Cutt
On Mon, Dec 22, 2008 at 6:15 AM, Joshua Tolley wrote:
> On Sun, Dec 21, 2008 at 10:25:59PM -0500, Rob
use the examine_variable() function in selfuncs.c except I would first
need a PlannerInfo and I don't think I can get that from inside the
join initialization code.
- Bryce Cutt
On Mon, Dec 15, 2008 at 8:51 PM, Robert Haas wrote:
> I have to admit that I haven't fully grokked what
Tom is correct that it
can be removed in the final version.
- Bryce Cutt
On Wed, Nov 5, 2008 at 7:22 AM, Joshua Tolley <[EMAIL PROTECTED]> wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> On Wed, Nov 5, 2008 at 8:20 AM, Tom Lane wrote:
>> Joshua Tolley wr
