Problems with cache

[mantczakaus]

Hi, I have reported before an issue with cache #30 In case of that one, the pipeline didn't cache the tasks even when it failed as far as pVAC. It was fixed with setting cache to 'lenient'. However, I'm still experiencing issues with cache when the pipeline fails before, e.g. on ReadBackedPhasing. When I was running the pipeline on WES data, it wasn't a big issue because the entire nextNEOpi pipeline took up to 10 hours but with over 100GB data it takes over a day to get to the ReadBackedPhasing task. Each time it fails, it goes back as far as SplitIntervals. Is there something else that could influence that behavior? Below a screenshot from Seqera Platform (grey = cached, green = succeeded, blue = running).