##Discussed##

  • setup the gateway using openvpn
    • will get the block of IP to be used as the subnet on virtual cluster
    • share the subnet on main cluster to be used as probable route
    • setup the gateway on main cluster using openvpn and public ip from kaizen
    • services like nfs on a node apart from controller on main cluster
    • check the reachibility
  • shouldn’t stop polling on main cluster
    • other nodes on the cluster might get affected
    • figure out an alternative using slurm reservation
  • reservation to mark the node status as maint
    • will not allocate the job when node is in maint status
    • node may not be marked as down when the VM is suspended
    • check the behaviour
    • can update the reservation duration as required to maintain the status
  • cvmfs file system for osg jobs
    • uses autofs
    • loads files on request
    • requires to register the site and setup a proxy
    • try exposing it as sshfs/nfs
      • try sshfs on autofs and expose the sshfs mount point as nfs
    • check will work or not
  • slurm nodes should know about other nodes across networks
    • tries to know the status of other nodes in the cluster
    • if not found then marks the node as down
    • will have to configure route for the node from main cluster to virtual cluster so as all the nodes to each other are visible

##To-Do##

  • check the slurm behaviour anf job retention when a node is in maint state
  • try exposing the cvmfs using sshfs/nfs and check the behaviour