If slurm_load_job() reported that our parent job ID was invalid, return immediately rather than trying
to use the (null) job_info struct later in the function. (In our testing we see segfaults about 40% of the time when trying to run bulk jobs via DRMAA otherwise.) This may also be the same issue discussed in #27.
If slurm_load_job() reported that our parent job ID was invalid, return immediately rather than trying to use the (null) job_info struct later in the function. (In our testing we see segfaults about 40% of the time when trying to run bulk jobs via DRMAA otherwise.) This may also be the same issue discussed in #27.