microsoft / microsoft/AzureTRE
AML Compute does not build when private in the workspace
Open
Nobody has claimed this yet.
bug
storysize/M
- Dominant language
- Python
- Stars
- 235
- Forks
- 192
- Avg merge
- 1d 23h
- Merged PRs (30d)
- 13
Description
The standard mcr docker images used for the AML compute clusters do not work within the TRE as access to conda / pypi repos are blocked and enabled via the package mirrors.
The setup for the nexus mirror needs to be baked in to the image used for compute in AML.
Plan is to:
- take a docker image
FROMthe standard one, and append the conda / pypi config to it. - build the docker image as part of the AML deploy
- push docker image to the AML ACR
- (optional) add an environment to AML which references the custom docker image
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by locating the AML deployment and the references to the standard MCR Docker image. Review how the deployment builds and publishes images, then configure the Nexus conda and PyPI mirrors, build the derived image, and push it to the AML ACR; confirm whether the optional AML environment is also required.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- azure, docker
- Domain
- cloud, devops, infrastructure, machine-learning
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100