| From: | Jeroen Vermeulen <jtvjtv(at)gmail(dot)com> |
|---|---|
| To: | daniel(at)yesql(dot)se |
| Cc: | Bharath Rupireddy <bharath(dot)rupireddyforpostgres(at)gmail(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: libpq: PQgetCopyData() and allocation overhead |
| Date: | 2023-03-02 19:44:48 |
| Message-ID: | CA+zULE5M8CujFba55g4NkGSSgRhxLeUhZbX+Q2oBivBcrmYyzQ@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
My apologies. The wiki said to discuss early, even before writing the code
if possible, but I added a link to the PR for those who really wanted to
see the details.
I'm attaching a diff now. This is not a patch, it's just a discussion
piece.
The problem was that PQgetCopyData loops use a lot of CPU time, and this
alternative reduces that by a lot.
Jeroen
On Thu, 2 Mar 2023 at 13:38, Daniel Gustafsson <daniel(at)yesql(dot)se> wrote:
> > On 1 Mar 2023, at 15:23, Jeroen Vermeulen <jtvjtv(at)gmail(dot)com> wrote:
>
> > PR for easy discussion: https://github.com/jtv/postgres/pull/1
>
> The process for discussing work on pgsql-hackers is to attach the patch to
> the
> email and discuss it inline in the thread. That way all versions of the
> patch
> as well as the discussion is archived and searchable.
>
> --
> Daniel Gustafsson
>
>
| Attachment | Content-Type | Size |
|---|---|---|
| PQhandleCopyData.patch | text/x-patch | 9.6 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Jeff Davis | 2023-03-02 19:45:51 | Re: Minimal logical decoding on standbys |
| Previous Message | reid.thompson | 2023-03-02 19:41:26 | Re: Add the ability to limit the amount of memory that can be allocated to backends. |