Re: [GSoC 2020] Questions About Performance Farm Benchmarks and Website

From: Mark Wong <mark(at)2ndQuadrant(dot)com>
To: Kalvin Eng <kalvin(dot)eng(at)ualberta(dot)ca>
Cc: pgsql-hackers(at)postgresql(dot)org
Subject: Re: [GSoC 2020] Questions About Performance Farm Benchmarks and Website
Date: 2020-03-03 16:54:29
Message-ID: 20200303165429.GA21925@2ndQuadrant.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi Kalvin,

On Tue, Mar 03, 2020 at 03:07:13AM -0700, Kalvin Eng wrote:
> Hi there,
>
> I am potentially interested in the performance farm project listed here:
> https://wiki.postgresql.org/wiki/GSoC_2020#Develop_Performance_Farm_Benchmarks_and_Website_.282020.29
>
> I've applied to the pgperffarm mailing list as well, but am waiting for
> moderator approval so I thought this list would be the best to ask about
> the performance farm code.
>
> Here are the questions based on the
> https://git.postgresql.org/gitweb/?p=pgperffarm.git;a=summary repo:
>
> - Why is a front end framework used instead of django templates?

I don't have a good answer for this, primarily because my knowledge on
the difference is weak...

> - Any reason why the server hasn't been containerized?

Simply because no effort has been put into it yet. Are you thinking for
ease of demoing or evaluating?

> - Django 1.11 will no longer be supported in April 2020, is it time to
> move to 2.2 LTS? (
> https://www.djangoproject.com/download/#supported-versions)

We want to match the same version the community infrastructure uses, so
yes, if that's the version they will be on.

> - What have been the issues with authentication integration to
> postgresql.org?

There is a custom authentication module that doesn't work outside of the
community infrastructure, and this project has been developed outside of
the community infrastructure. We haven't come up with a way to bridge
that gap yet.

> - Should the client be turned into a package for package managers (e.g.
> pypi, DPKG, brew, etc.)?

I think that would be a plus.

> - The project description mentions refactoring to Python 3, but it seems
> like that was completed last GSoC?

Yeah, I think that's been squared away...

> - Should the performance visualizations be added again?

Yes, that would good to have.

> I've also looked at past mailing lists for this project, but am interested
> in hearing current insights from the community:
>
> -
> https://www.postgresql-archive.org/GSoC-2019-report-amp-feedback-td6100606.html
> -
> https://www.postgresql-archive.org/GSoC-2019-Proposal-Develop-Performance-Farm-Database-and-Website-td6079058.html
> -
> https://www.postgresql-archive.org/GSoC-Summery-of-pg-performance-farm-td6034578.html
> -
> https://www.postgresql-archive.org/GSOC-18-Performance-Farm-Project-Initialization-Project-td6010380.html
> -
> https://www.postgresql-archive.org/GSOC-18-Performance-Farm-Project-td6008120.html
> - https://www.postgresql-archive.org/performance-test-farm-td4388584.html

Regards,
Mark

--
Mark Wong
2ndQuadrant - PostgreSQL Solutions for the Enterprise
https://www.2ndQuadrant.com/

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2020-03-03 17:10:15 Is it time to retire type "opaque"?
Previous Message Ashutosh Bapat 2020-03-03 16:48:40 Re: [HACKERS] advanced partition matching algorithm for partition-wise join