I am trying to write the cluster config for LSF and I noticed that the generated cluster_job_id is not as expected (it corresponds to the queue name). Any hint?
Thanks for reporting this- this is currently a bug in cryoSPARC, where the scheduler only gets the job id from the last value of the output returned by submitting a job. This works for SLURM, but not most other cluster schedulers. A fix for this has been made, and will be released in the next version of cryoSPARC (v2.16.3+)
I am using v3.2.0+210817 and this doesn’t seem to be fixed. Our cluster is UGE and we use qsub commands, and the Cluster Job ID reports “submitted” rather than a number.
Like in this solution, Installation on SGE cluster, I’ve fixed it using the -terse option in the qsub command, and now we have a real Cluster Job ID and the full info returned by qstat.
Thank you for the update.