Roadmap for bookworm migration

[gbip]

Hi,

Are there any plan to migrate from bullseye to bookworm ? How can we help ?

I am interested in using GEOS 3.11 features (for example ST_SimplifyPolygonHull), which is the version of GEOS available on debian bookworm.

Debian bullseye only has GEOS 3.9.0-1.

Best regards,

[ImreSamu]

Are there any plan to migrate from bullseye to bookworm ?

I've been considering this for a while, but it's not that simple.

There are two expected issues with the new Debian version:
a.) libseccomp

• Docker image for Postgres 14 based on BookWarm is broken somehow docker-library/postgres#1100

b.) glibc + collation version mismatch

• Docker image for version 15.3 throws collation warnings docker-library/postgres#1099

Because of these issues, some people would prefer to stay with the bullseye version, but currently the postgis/docker-postgis repository only supports one Debian version.

The ideal solution would be to support both Debian versions, for example:

• postgis/postgis:15-3.3-bullseye
• postgis/postgis:15-3.3-bookworm

But this would require significant restructuring, and I won't be able to start working on it in the next two weeks due to other commitments.
But if it's urgent and someone makes a PR by then, I'd be happy to help with testing.
If someone starts working on it, please indicate so here in a comment.

[ImreSamu]

@gbip :

I wanted to let you know that I've started working on supporting multi-debian (bookworm, bullseye) and multi-alpine postgis images.

I hope to provide a minimal Pull Request soon for testing. However, I cannot commit to a specific deadline yet.

[RangaSamudrala]

Are there any plan to migrate from bullseye to bookworm ?

I've been considering this for a while, but it's not that simple.

There are two expected issues with the new Debian version: a.) libseccomp

• Docker image for Postgres 14 based on BookWarm is broken somehow docker-library/postgres#1100

b.) glibc + collation version mismatch

• Docker image for version 15.3 throws collation warnings docker-library/postgres#1099

Because of these issues, some people would prefer to stay with the bullseye version, but currently the postgis/docker-postgis repository only supports one Debian version.

The ideal solution would be to support both Debian versions, for example:

• postgis/postgis:15-3.3-bullseye
• postgis/postgis:15-3.3-bookworm

But this would require significant restructuring, and I won't be able to start working on it in the next two weeks due to other commitments. But if it's urgent and someone makes a PR by then, I'd be happy to help with testing. If someone starts working on it, please indicate so here in a comment.

Will there be effort to support Postgis v12 using Bookworm?

[ImreSamu]

@RangaSamudrala

Will there be effort to support Postgis v12 using Bookworm?

I'd definitely like to!

• 12-3.4-bookworm ✔️
  • https://registry.hub.docker.com/r/imresamu/postgis/tags?page=1&name=12-3.4-bookworm

Development environment ( not stable ) - but you can check the plan :

• https://hub.docker.com/r/imresamu/postgis
• code: https://github.com/ImreSamu/docker-postgis

PR: #356

[zachsa]

Hi,

I am keen on this (I like the postgis docker image, and want to extend it with another extension that has pre-built binaries for bookworm)

[ardentperf]

FYI - quick reminder that users can't safely switch from bullseye images to bookworm images without logical pg_dump-and-load or logical replication. Switching the container image from bullseye to bookworm and keeping the same storage (whether via PVCs or StatefulSets) can corrupt your database.

Probably worth thinking about how this will work end-to-end.

The cause is the stability issues with linguistic collation (especially in libc) and a number of database corruptions have happened in recent years related to this. Here's a talk from the postgres development conference earlier this year on the topic https://www.youtube.com/watch?v=KTA6oau7tl8

This is the same issue referenced by @ImreSamu above

Supporting both bullseye and bookwork (similar to the base postgres image) should be fine

[lnicola]

While that's true, in most cases it's not real corruption, and you can fix it by refreshing the collation and/or rebuilding the affected indexes.

More importantly, it applies just as well to a native (distro package) installation and to the base postgres image. Because of that, I don't think the postgis image should "gatekeep" bookworm. Bullseye is EOL, and this decision will only make it harder for users when trixie comes out.

[haubourg]

Hi here,
I am facing the very same issue of a too large gap between too old GEOS in bookworm and bleeding edge in alpine. I won't risk the switch from debian to alpine based image on my critical and massive production database ( ref https://nickjanetakis.com/blog/benchmarking-debian-vs-alpine-as-a-base-docker-image ) .

Is there any work planned here when Trixie will be released months or should I plan building my own images ?

Thanks a lot for the updates