| From: | Alexander Korotkov <aekorotkov(at)gmail(dot)com> |
|---|---|
| To: | Heikki Linnakangas <heikki(dot)linnakangas(at)enterprisedb(dot)com> |
| Cc: | pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Double sorting split patch |
| Date: | 2011-09-21 21:56:25 |
| Message-ID: | CAPpHfds1YZSg6tdUrX+bFS3q=NFN+sSVx_awvbb0hKupUrmQ-g@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
I've processed the results of the tests with double sorting split which I've
sheduled for buffering build. I've updated wiki page with them:
http://wiki.postgresql.org/wiki/Fast_GiST_index_build_GSoC_2011#Testing_results
Raw results of query speed measues are in the attachment. There number of
pages accesse is presented in dependency of table, buffering build and split
method.
Note, that tests were run with not last version of fast build patch
("gist_fast_build-heikki-0.14.1.1.patch" was used). Therefore, build time
with buffering can be better.
------
With best regards,
Alexander Korotkov.
| Attachment | Content-Type | Size |
|---|---|---|
| search_time.txt | text/plain | 10.7 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Josh Berkus | 2011-09-21 23:52:35 | citext operator precedence fix |
| Previous Message | Tom Lane | 2011-09-21 21:49:21 | Re: EXPLAIN and nfiltered, take two |