Inspecting the apercal.log's for 190326042 there is output that is still being written there for other datasets including 190327042 and 190328001. This mix up appears to be continuing with individual beams as well.
From Tammo Jan: Argh. I could imagine that the logging doesn't get cleaned up between runs of the pipeline. In start_pipeline there is an attempt to remove output of parallel loggers but also that is a bit broken.
Inspecting the apercal.log's for 190326042 there is output that is still being written there for other datasets including 190327042 and 190328001. This mix up appears to be continuing with individual beams as well. From Tammo Jan: Argh. I could imagine that the logging doesn't get cleaned up between runs of the pipeline. In start_pipeline there is an attempt to remove output of parallel loggers but also that is a bit broken.