Open
Conversation
This should be the general strategy for collecting input and output files for ARC, DIRAC, AWS batch etc.
Change base image from `conda/miniconda3` (based off Debian Stretch) to `python:3.12-bookworm`. Miniconda is not required in the base image. Add the Galaxy Depot repository, which provides SLURM DRMAA packages for Debian Buster and newer releases. Do not install the package `apt-transport-https`, it is now a dummy package, see https://packages.debian.org/en/bookworm/apt-transport-https. Install the package `slurm` instead of `slurm-llnl`. Newer versions of the `munge` package include the binary `/usr/sbin/mungekey` instead of `/usr/sbin/create-munge-key`. Nevertheless, the key seems to be created automatically when installing the package, as running `mungekey` yields 'mungekey: Error: Failed to create "/etc/munge/munge.key": File exists'.
Build wheel automatically when building the Docker image. Exclude the source code from the output image through a multistage build.
…oexecutionLaunchMixin` to `BaseRemoteConfiguredJobClient`
4 tasks
jmchilton
reviewed
Jul 3, 2025
jmchilton
reviewed
Jul 3, 2025
Contributor
Author
jmchilton
reviewed
Jul 3, 2025
…Collector` At the moment, JSON staging and outputs manifests are constructed by tracking all actions mapped by the `FileActionMapper` using a list `FileActionMapper.actions`. This makes the `FileActionMapper` stateful, requires including `file_type` as keyword argument for `BaseAction` and its children, requires defining a finalize()` method for `FileActionMapper` and for `BaseAction` and its children. Paying the small price of refactoring `JsonTransferAction`, generate the staging manifest from `FileStager.transfer_tracker.remote_staging_actions` and the output manifest as `ResultsCollector` collects the outputs.
Set `JsonTransferAction.whole_directory_transfer_supported` to `False`, as the job files API is not capable of serving directories.
Using `basename(action.path)` creates a flat structure for each file type (e.g. job_directory/unstructured/human.fa), but Pulsar expects tree structures to work (e.g. job_directory/unstructured/f0d0164494db6cbf92c12aeb6119ac38/bwa/human.fa).
Implement a Pulsar client that runs jobs on computing infrastructure using the Advanced Resource Connector (ARC) middleware.
This was referenced Jul 17, 2025
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Implement a Pulsar client that runs jobs on computing infrastructure using the Advanced Resource Connector (ARC) middleware.
Requires #399, jmchilton/simple-job-files#1, jmchilton/simple-job-files#2, jmchilton/simple-job-files#3.