Hi,
04.02.2022 03:52, Alexander Korotkov wrote:
Thank you for the revision. This patch looks good to me. I've
slightly adjusted comments and formatting and wrote the commit
message.
I'm going to push this if no objections.
While exploring the gist test coverage (that is discussed in [1])
Hi!
On Wed, Jan 26, 2022 at 7:07 PM sergei sh. wrote:
> I've fixed issues 2 -- 4 in attached version.
>
> GUC parameter has been removed for the number of pages to collect
> before splitting and fixed value of 4 is used instead, as discussed
> off-list with Andrey Borodin, Aliaksandr Kalenik,
Hi,
On 2022-01-18 23:26:05 +0300, sergei sh. wrote:
> I've addressed Andrey Borodin's concerns about v2 of this patch by
> Aliaksandr
> Kalenik in attached version. Change list:
> * Number of pages to collect moved to GUC parameter
> "gist_sorted_build_page_buffer_size".
> *
> 19 янв. 2022 г., в 09:31, Andrey Borodin написал(а):
>>
>> I've addressed Andrey Borodin's concerns about v2 of this patch by Aliaksandr
>> Kalenik in attached version.
>
> Thank you! I'll make a new iteration of review. From a first glance
> everything looks good, but
> 19 янв. 2022 г., в 01:26, sergei sh. написал(а):
>
> Hi,
>
> I've addressed Andrey Borodin's concerns about v2 of this patch by Aliaksandr
> Kalenik in attached version.
Thank you! I'll make a new iteration of review. From a first glance everything
looks good, but
Hello hackers,
On Tue, Jan 18, 2022 at 11:26 PM sergei sh. wrote:
> Hi,
>
> I've addressed Andrey Borodin's concerns about v2 of this patch by
> Aliaksandr
> Kalenik in attached version.
>
[snip]
This patchset got some attention in the PostGIS development channel, as it
is important to really
Hi,
I've addressed Andrey Borodin's concerns about v2 of this patch by
Aliaksandr
Kalenik in attached version. Change list:
* Number of pages to collect moved to GUC parameter
"gist_sorted_build_page_buffer_size".
* GistSortedBuildPageState type renamed to GistSortedBuildLevelState.
*
> 18 янв. 2022 г., в 03:54, Björn Harrtell
> написал(а):
>
> There might be some deep reason in the architecture that I'm unaware of that
> could make it difficult to affect the node size but regardless, I believe
> there could be a substantial win if node size could be controlled.
That's
Hi Aliaksandr,
Nice work on this. I've been following it a bit since the regression when
it was noted and it sparked renewed interest in R-tree structure and
optimization for me.
As for ideas. I'm not deep into details of postgresql and gist, but I've
learned that the node size for gist indexes
Hi,
here are some benchmark results for GiST patch:
https://gist.github.com/mngr777/88ae200c9c30ba5656583d92e8d2cf9e
Code used for benchmarking: https://github.com/mngr777/pg_index_bm2,
see README for the list of test queries.
Results show query performance close to indexes built with no
After further testing, here is v2 where the issue that rightlink can be set
when an index page is already flushed is fixed.
On Sat, Dec 25, 2021 at 4:35 PM Andrey Borodin wrote:
>
> Hi Aliaksandr!
>
> Thanks for working on this!
>
> > Benchmark summary:
> >
> > create index roads_rdr_idx on
Hi Aliaksandr!
Thanks for working on this!
> Benchmark summary:
>
> create index roads_rdr_idx on roads_rdr using gist (geom);
>
> with sort support before patch / CREATE INDEX 76.709 ms
>
> with sort support after patch / CREATE INDEX 225.238 ms
>
> without sort support / CREATE INDEX
Hey!
Postgres 14 introduces an option to create a GiST index using a sort
method. It allows to create indexes much faster but as it had been
mentioned in sort support patch discussion the faster build performance
comes at cost of higher degree of overlap between pages than for indexes
built with
13 matches
Mail list logo
