I found a new error while running with some SRA data:

Command exit status:
  255

Command output:
  (empty)

Command error:

  WARNNIG: different read numbers of the 446823 pack
  Read1 pack size: 101
  Read2 pack size: 256

  ERROR: input files don't contain identical amount of reads

The module is FASTP. What's the fastest safe way to solve this? I'm happy to discard a few bad reads as long as the rest of the file stays intact

I used seqkit in the past

Is there a possibility that the download from SRA got corrupted? Because it would be surprising if the data is supposed to be like that.

Perhaps, although fetchngs ran without errors

If the file system does not report an error (I've had this happen in cases of write errors due to overload), the pipeline would never know, and since the files are not actually being used by fetchngs, you only see the errors in the next pipeline and blame that. (In my case I saw the result of file truncation in the same pipeline, but in a downstream process, whereas the upstream process producing the faulty file had exit code

0

.) In short, if there are check sums (MD5, etc.), I would verify those, otherwise just retry the download and see of the files are different. 🤞