After reading the logic of removing useless join, I think the comment of
this might need to be changed: "Currently, join_is_removable only succeeds
if sjinfo's right hand is a single baserel. " could be changed to
"Currently, join_is_removable only succeeds if sjinfo's min_righthand is a
single bas
e provided in this paper also has some defects, as
mentioned above, the scheme relies on the collected samples, which will
lead to a significant increase in the storage overhead of statistical
information.
I'd like to hear your opinions.
ywgrit.
ywgrit 于2023年12月22日周五 16:20写道：
> The possib
col_1, pred_col_2,
... pred_col_n) with where clause could be able to estimated accurately.
I'd like to hear your opinions.
Regards.
ywgrit.
Tomas Vondra 于2023年12月18日周一 20:53写道：
>
>
> On 12/18/23 11:40, Richard Guo wrote:
> >
> > On Mon, Dec 18, 2023 at 7:31 AM To
uld suggest to do the detection when the
index is created, because then we can get the details of the index and give
a warning in the way you mentioned.
Tom Lane 于2023年12月13日周三 23:01写道：
> ywgrit writes:
> > I forbid to create indexes on whole-row expression in the following
> patch.
&
/flat/e48a5d9a2d3d72985d61ee254314f5f5f5444a55.ca...@cybertec.at
I forbid to create indexes on whole-row expression in the following patch.
I'd like to hear your opinions.
--
Best Wishes,
ywgrit
diff --git a/src/backend/commands/indexcmds.c b/src/backend/commands/indexcmds.c
index cd23ab3b25..e4451b1d36 100644
--- a/src/backen
ch made me
wonder how data field is copied into the HTAB?
But at the time I ignored a note above: "Caller is expected to fill the
data field on return". Now I know that the data field needs to be filled
manually, so it was my misuse. Thanks for the correction!
Thanks
Tom Lane 于2023年9月1
The definition of hashctl is shown below
typedef struct HASHCTL
{
long num_partitions; /* # partitions (must be power of 2) */
long ssize; /* segment size */
long dsize; /* (initial) directory size */
long
