fix(container): rebuild ccuriqueo/vi-python:3.8 to include procps

[juanu]

Bug

ABUNDANCE:GET_CIGAR (and potentially other processes using ccuriqueo/vi-python:3.8) fails with exit code 1 even though the Python script completes successfully.

Root cause: Nextflow's nxf_trace_linux() function — injected into every task's .command.run — requires the ps command to collect per-task CPU/memory metrics. The published image ccuriqueo/vi-python:3.8 was built before procps was added to the Dockerfile, so ps is absent and the tracer calls exit 1 immediately.

The relevant line in .command.run:

command -v ps &>/dev/null || { >&2 echo "Command 'ps' required by nextflow to collect task metrics cannot be found"; exit 1; }

Current workaround

The host /usr/bin/ps binary is bind-mounted read-only into the container via docker.runOptions in nextflow.config:

docker.runOptions = '-u $(id -u):$(id -g) -v /usr/bin/ps:/usr/bin/ps:ro'

This is fragile — it depends on the host ps being ABI-compatible with the container's libc.

Permanent fix

The containers/Dockerfile already includes procps. The image just needs to be rebuilt and pushed:

docker build -t ccuriqueo/vi-python:3.8 containers/
docker push ccuriqueo/vi-python:3.8

After pushing:

• Remove the -v /usr/bin/ps:/usr/bin/ps:ro bind-mount from docker.runOptions in nextflow.config (both docker and standard profiles)
• Verify: docker run --rm ccuriqueo/vi-python:3.8 which ps
• Re-run the pipeline to confirm ABUNDANCE:GET_CIGAR completes successfully