Hi Everyone! I am running nf-core/rnaseq with the test_full pipeline. I want to use my Cloudflare R2 bucket for workdir and outdir. However I believe I am getting this error because it wants to get input files from an AWS S3 bucket and tries to look for them in my R2 bucket, where they don’t exist? Is there a way to run this without having to stage all the input data from AWS S3 into my R2 bucket beforehand?

This is a good question. The problem is that R2 is a S3-compatible object storage, but that the

s3://

prefix will either be routed to AWS or to Cloudflare, depending on your aws scope config. So either it tries to write the output files to a non-existing AWS bucket, or it doesn’t find the input files. On the Nextflow Slack, there is an AWS channel, so maybe they know more or you find a rarely used parameter in the AWS config section in the docs. Maybe you can specify either bucket with some absolute URL instead of just the

s3://

prefix?