| From: | Claudio Freire <klaussfreire(at)gmail(dot)com> |
|---|---|
| To: | Leonardo Francalanci <m_lists(at)yahoo(dot)it> |
| Cc: | PostgreSQL-Dev <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Fast insertion indexes: why no developments |
| Date: | 2013-11-05 17:58:11 |
| Message-ID: | CAGTBQpYWRHAWfHtTxpd-_gzMFgbvVLcmmmr4fw0+MF=wc=FFEg@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Tue, Nov 5, 2013 at 2:52 PM, Leonardo Francalanci <m_lists(at)yahoo(dot)it> wrote:
> Jeff Janes wrote
>> Some experiments I did a few years ago showed that applying sorts to the
>> data to be inserted could be helpful even when the sort batch size was as
>> small as one tuple per 5 pages of existing index. Maybe even less.
>
> Cool!!! Do you have any idea/hint on how I could try and replicate that?
> Do you remember how you did it?
I do it regularly by sorting tuples before inserting/updating. It
helps quite significantly for batches of ~1000 tuples (well, in my
case).
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Josh Berkus | 2013-11-05 18:00:15 | Re: List of "binary-compatible" data types |
| Previous Message | Leonardo Francalanci | 2013-11-05 17:52:40 | Re: Fast insertion indexes: why no developments |