https://github.com/zonca/docker-jupyter-cdms-light
Docker CDMS image for JupyterHub to be used together with the CDMS software stack on CVMFS
https://github.com/zonca/docker-jupyter-cdms-light
Last synced: 10 months ago
JSON representation
Docker CDMS image for JupyterHub to be used together with the CDMS software stack on CVMFS
- Host: GitHub
- URL: https://github.com/zonca/docker-jupyter-cdms-light
- Owner: zonca
- License: bsd-2-clause
- Created: 2020-03-06T16:51:04.000Z (almost 6 years ago)
- Default Branch: master
- Last Pushed: 2022-10-13T03:12:06.000Z (about 3 years ago)
- Last Synced: 2025-01-13T10:30:32.920Z (11 months ago)
- Language: Dockerfile
- Size: 29.3 KB
- Stars: 0
- Watchers: 2
- Forks: 1
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# docker-jupyter-cdms-light
Docker CDMS image for JupyterHub to be used together with the CDMS software stack on CVMFS
First rebuilt the Jupyter Docker stacks out of Centos 7 instead of Ubuntu,
see:
I build this manually (later will possibly setup autobuild) and push to Docker Hub:
We are interested in the image tagged `tensorflow`.
Finally we inherit from that image to create `docker-jupyter-cdms-light`:
As examples, checkout the `Dockerfile` files in
## Versioning

Generally we update a few times a year and we do 3/4 releases the same day, so let's use same versioning system used by Ubuntu:
`YY.0M.MICRO`
See
Better to not use `latest` in production, always make a tag and use that. See [the Image build section](#image-build) for instructions on how to do this.
## Image build
The container is hosted [on DockerHub](https://hub.docker.com/r/zonca/docker-jupyter-cdms-light)
Unfortunately [Docker stopped providing free autobuilds](https://www.docker.com/blog/changes-to-docker-hub-autobuilds),
so now we need to build a new image on a machine with Docker and then push to Dockerhub.
The `Makefile` in this directory provides useful docker commands. All commands assume (1) that you set a `VERSION` environment variable for the tag (see [the versioning section](#Versioning)) and (2) that you can run `docker` without `sudo`. See [the Docker documentation](https://docs.docker.com/engine/install/linux-postinstall/) if you need to set this up.
* `make build` builds a container from the Dockerfile
* `make shell` will drop you into an image shell; useful if you want to check anything before pushing
* `make push` uploads the container to [dockerhub](https://hub.docker.com/r/zonca/docker-jupyter-cdms-light). For this command to work you will need access to this dockerhub project - get a dockerhub account if you don't already have one and request access from the maintainer of this repository. You can `make login` if you are not already authenticated to dockerhub.
If your shell supports tab completion, `make` spacetabtab will show all available commands (or "targets" in make-language). You can also open `Makefile` in a text editor to see available targets.
## Deploy in production
The version of the image used in production is defined in `config_standard_storage.yaml`:
```yaml
image:
name: "zonca/docker-jupyter-cdms-light"
tag: "2020.11.25"
```
need to update that and then run `install_jhub.sh`.