Just re-discovered that ampliseq fails on AWS with

Essential container in task exited

for 2.16.1, 2.16.0, 2.15.0 but (I think Rike) managed to run 2.15.0 on the second try. Similar problem was for 2.13.0 and it seemed related to fusion (just parroting here). Is there something I can do that this will work in future right away?

I turned off fusion for QIIME processes. There is not much we can do, aside from setting scratch = true in the configuration on the aws submission for all QIIME related processes. Are these all old runs or are there new runs that are also affected? There have been quite a few releases that should address some of these issues

2.13.0: April 2025, 2.15.0: Oct. 2025, 2.16.0: 16.01.26, 2.16.1: 24.01.26 So over a large time span, rather recent.

Hey! When you resubmit the run on platform you can configure just as you wuld iwht a local config file . Resubmitted this one now: https://cloud.seqera.io/orgs/nf-core/workspaces/AWSmegatests/watch/aItIm07lfO232/v2/logs

Failed unfortunately