The location of your configuration and log files will be determined by your cluster management decisions. We typically place files required at runtime on the sss1 (the main administrative node) in /cplant/vm/VM-NAME/nfs-cplant/etc which we find at /cplant/etc on all the service and compute nodes. (VM-NAME of course would be replaced with the name of your virtual machine. Then the file system at /cplant/vm/VM-NAME/nfs-cplant is mounted at /cplant on the all service and compute nodes in virtual machine VM-NAME.)
The /cplant/etc file system is read only. We need a writable directory for the user log files written by yod and bebopd, for the restart file written by the bebopd, and for the log files written by the PBS daemons.
We have had trouble allowing yod processes on different service nodes to write the same NFS mounted log file. This is in spite of the fact that we wrote our own file locking routine in yod that should work in the environment in which we were running it. For this reason, we prefer to set up the user log file so that there is a separate one for each service node, either through links to a separate file for each service node or by writing the user log file to a local disk on the service node.
We find installing a local disk in the service node and placing log files and restart files there works best for us.