Closed KrisThielemans closed 3 months ago
docker run --rm -it -p 9999:8888 --gpus all ghcr.io/synerbi/sirf:jupyter-devel
fails as well. We could presumably say that we don't push the devel
image anywhere, which probably reduces confusion. But then an easy line in the doc how to build it would be highly appreciated.
I see that a PR merge triggers docker builds, where edge
images are pushed (somewhere).
https://github.com/SyneRBI/SIRF-SuperBuild/actions/runs/8240707009/job/22536603863#step:11:146
The table at the top of #718 describes the relation between git repo tags and docker image tags.
docker tag SuperBuild branch/tag latest
,latest-gpu
latest tag vM.m.p
M
,M.m
,M.m.p
,M-gpu
,M.m-gpu
,M.m.p-gpu
tag vM.m.p
edge
,edge-gpu
master
only build & test (no tag) CI (current commit) devel
,devel-gpu
master
withcmake -DDEVEL_BUILD=ON -DBUILD_CIL=ON
I think (similar to https://github.com/TomographicImaging/CIL#docker) we should document this in the README.
TL;DR you need to push a git-tagged release :)
I think (similar to https://github.com/TomographicImaging/CIL#docker) we should document this in the README.
sure
TL;DR you need to push a git-tagged release :)
"you" = "me"? Doesn't it happen via GitHub Actions?
Yes GHA will do it when a tag is pushed.
By the way, docker Hub contains edgegpu
, not edge-gpu
oh, and there's no jupyter
in the above table :-)
By the way, docker Hub contains
edgegpu
, notedge-gpu
fixed in #891
oh, and there's no
jupyter
in the above table :-)
Hmm gewd point. A bit conflicted on this. Discussion in #891
I have no idea which tags are available, and it seems ghcr.io doesn't have a search/list facility (or not an obvious one anyway).
The README mentions
jupyter
andjupyter-gpu
, but there is more.I see
core
(but maybe this is internal and doesn't need to be documented) https://github.com/SyneRBI/SIRF-SuperBuild/blob/fd4e9720965f846889e468eb938c88e5805ff5d3/.github/workflows/docker.yml#L81 although the tag might be different (a bug?) https://github.com/SyneRBI/SIRF-SuperBuild/blob/fd4e9720965f846889e468eb938c88e5805ff5d3/.github/workflows/docker.yml#L94-L95There's a 2 push statements https://github.com/SyneRBI/SIRF-SuperBuild/blob/fd4e9720965f846889e468eb938c88e5805ff5d3/.github/workflows/docker.yml#L107 https://github.com/SyneRBI/SIRF-SuperBuild/blob/fd4e9720965f846889e468eb938c88e5805ff5d3/.github/workflows/docker.yml#L112
I see
jupyter-devel
being created https://github.com/SyneRBI/SIRF-SuperBuild/blob/fd4e9720965f846889e468eb938c88e5805ff5d3/.github/workflows/docker.yml#L176 But I cannot find apush
for thedevel
image.Indeed, trying to create a Codespace using
jupyter-devel
fails https://github.com/SyneRBI/SIRF-Exercises/blob/devel-container/.devcontainer/devcontainer.jsonhttps://github.com/SyneRBI/SIRF-SuperBuild/tree/master/docker#docker-stacks mentions
synerbi/jupyter:sirf
but alsosynerbi/sirf:latest
, which isn't documented anywhere?By the way, https://hub.docker.com/r/synerbi/sirf/tags? gives me mostly stuff that is 4 months old, aside from
edge
andedgegpu
. I have no clue what those are.Looks like, we have to update our
CHANGES.md
to say that we no longer push to DockerHub and people need to read new instructions.