preload-size error exits with code 0

[jvolkening]

If more memory is requested using --preload-size than is available, KrakenUniq throws an error and terminates (ERROR: Could not allocated memory for loading database chunks). However, the exit code in bash is still 0. I'm not sure why this is because the next line of code in the source is exit(-1);.

The issue can be reproduced simply by requesting more memory than the machine has, e.g on a 128GB machine:

krakenuniq --db db --report-file report.txt --output results.tsv --preload-
size 150G in.fq && echo $?

This causes problems in our Nextflow pipelines because the error situation is not detected.