Hi.

I was looking for a way to access cluster_config dictionnary to define specific ressources in some particular rules (like -Xmx memory for java programs).

I understood that it's not ideally, because the Snakefile became dependant of the execution mode on a cluster.

I tried to understand the resources options and rule's tag. But it's not clear for me how it works and how it interact with --drmaa syntaxe or with --cluster

For me, ideally, I would like to define a resources.yaml that can be accessible in the Snakefile and in the submit command sheme (something like : " qsub -pe parallel_smp {resources.cpu} -l mem={resources.mem}" )

Is that possible ? If not, what is the solution : - define resources in rule AND in cluster-config if executing Snakefile on a cluster ?

I don't see any other solutions for now ....but I believe that we can do better, no ?

Any help would be much appreciate.

Best regards

Maria