--disk-limit and --disk-limit-tmp => limits not enforced

[nick-youngblut]

When I set fasterq-dump --disk-limit 0.1GB --disk-limit-tmp 0.1GB, the limits are not enforced. The fasterq-dump jobs never throw an error, regardless of the (temp) output file sizes.

Also, the docs at https://github.com/ncbi/sra-tools/wiki/HowTo:-fasterq-dump mention --disk-limit and --disk-limit-tmp, but do not actually explain what they are. The fasterq-dump CLI docs also do not explain the --disk-limit and --disk-limit-tmp parameters. For example, one could assume multiple possibilities on what happens when a disk limit is reached:

1. fasterq-dump will exit gracefully when the limit is reached, and only the reads already downloaded are written to the file system
2. faster-dump just throws a non-zero exit

I believe the the 2nd option is the actual behavior. The 1st option would be much more desirable, especially for compute environments with usually small, defined disk limits (e.g., cloud batch VMs). The 1st option would then act somewhat like --maxSpotId from fastq-dump, which is lacking in fasterq-dump.