See "alloc_node" in the job_descriptor.
This identifies the submit node name, but not a cluster (for example,
the job could be submitted from someone's desktop with some
configurations).
There are a bunch of locks set when the job_submit plugin is called,
so you shouldn't execute any slurm commands and use most APIs, but the
functions in src/common can be used.
Quoting Martins Innus <[email protected]>:
Hello,
What is the best way to get the cluster that the job was
submitted to from a job submit plugin? I didn't see it in the job
descriptor struct and I wasn't sure what API's are allowed to be
called from this context.
Thanks
Martins
