When resuming our pipelines that have been launched from Seqera Tower, either via the LaunchPad, or programmatically via an Action, the resume function works inconsistently. When using -resume on the same pipelines launched locally, resume works every time. To debug this, I’d like to examine the cache hashes.
When running a Nextflow pipeline from the CLI I can add -dump-hashes.
Where and how can I add this argument in the Seqera Tower interface to the same effect?
Found it in the docs here: https://docs.seqera.io/platform/23.2.0/faqs#nextflow-configuration
" As of Nextflow v22.09.1-edge, when invoking a pipeline from Tower, you can specify Nextflow CLI run arguments by setting the NXF_CLI_OPTS environment variable via pre-run script:
export NXF_CLI_OPTS='-dump-hashes'
Sadly, on implementation, this didn’t work 
Relaunching with:
Led to…
This was with N E X T F L O W ~ version 24.04.2
You can enable this in the Nextflow configuration section using:
dumpHashes = true
It’s a little hard to find because it’s under Configuration — Nextflow documentation