Most of the time, I'm able to infer what the fields returned by various queries mean. However, there are times when things are a little bit ambiguous.
For example, I'm currently trying to put together some custom cluster statistics. As part of this, I'm trying to characterize summary utilization statistics for CPUs on a per cluster basis. When I query pyslurm.node().get_node('node-id') I find two interesting values, 'alloc_cpus' and 'cpu_load'. From this I would assume cpu_load is some kind of aggregate percentage load for the whole node, and alloc_cpus is the number of CPUs that have been spoken for by current jobs assigned to the node. However, that's a guess. Is there any exhaustive definition for what these fields represent? Are they directly taken from the C API? If so, is there a reference somewhere for that C API that I can examine in detail? I've googled around quite a bit and have as yet not been able to find an API doc short of spelunking through the source itself.
If anyone knows of such a reference, that would be extremely helpful.
Many thanks!