1. 18 May, 2018 2 commits
    • Neil McGlohon's avatar
      Dragonfly Plus: Intermediate and Rerouting confs · c5d0b5a1
      Neil McGlohon authored
      This commit adds functionality to the dragonfly plus model. The
      functionality is to make whether source group rerouting and
      intermediate router choice a configurable option that is set
      in the config file.
      To use said configurables, use these config variables in your
      config file:
      	source_group_rerouting= {"off" | "on"} default="on";
      	intermediate_choice= {"leaf" | "spine" | "both"} default="leaf";
      If source group rerouting is on, then if a packet is currently being
      routed by a spine router in the source group, but the spine router
      does not have a connection to the intermediate group, then the spine
      will route the packet to another spine router within the source
      group who does have said connection to the intermediate group.
      If source group rerouting if off, then if a packet being currently routed
      by a spine router in the source group cannot be directly routed to the
      pre specified intermediate group in the message, then a new intermediate
      router/group will be chosen from the list of possible intermediate
      groups/routers that the source spine has connections to. It is not
      rerouted within the source group.
      Intermediate choice determines what type of routers can be chosen
      as an intermediate router. If "leaf" is set, then intermediate routers
      will only be of type leafs. Similar for "spine". If "both" is selected
      then it could be either.
      Additional features added in this commit:
      New configurable:
      	notification_on_hops_greater_than={"<int>"} default = MAX_INT
      This new configurable sets the maximum number of hops allowed before a
      printed notification of the number of hops that a current packet
      is at is output to stdout. This is to help debug and know when routing
      decisions are producing unexpected results.
    • Neil McGlohon's avatar
      Add Dragonfly Custom Print Config · 5f62907b
      Neil McGlohon authored
  2. 17 May, 2018 1 commit
  3. 16 May, 2018 1 commit
  4. 15 May, 2018 5 commits
  5. 14 May, 2018 4 commits
  6. 11 May, 2018 11 commits
  7. 10 May, 2018 2 commits
  8. 09 May, 2018 5 commits
  9. 04 May, 2018 2 commits
  10. 03 May, 2018 2 commits
  11. 01 May, 2018 2 commits
  12. 30 Apr, 2018 3 commits
    • Neil McGlohon's avatar
      hotfix: comment out a print statement · 07be9b5e
      Neil McGlohon authored
    • Neil McGlohon's avatar
      ModelNet MPI Replay: Rand Perm Traffic Fix · daaf1739
      Neil McGlohon authored
      This commit changes the behavior of the random permuation traffic.
      My understanding is that s->gen_data is what keeps track of how much
      data has been generated by the LP where syn_data keeps track of how much
      data has been received by the LP. Since the random permutation traffic
      pattern cares about how much data has been _generated_, the use of
      syn_data in the gen_synthetic_tr() method's PERMUTATION case has been
      changed to gen_data.
      Additionally, the previous implementation of the prev_switch, as far
      as I could tell, was designed to help keep track of how much data had
      been generated by the LP since the last destination change. It didn't
      appear to do that. I've made a change that updates this field to be
      the amount of data generated at the time of the last switch, therefore
      the difference between gen_data and this field will equal the amount
      of data generated since last switch which is what we want to compare
      against the permuation threshold.
      In order to accomodate this behavior, I've had to add one more unsigned
      long into the nw_message ROSS message struct. The message size here is
      now 640 (changed from 624).
      Additionally, this source file hadn't yet been able to accomodate
      synthetic5 traffic option. It maxed out at 4. This fix was included
      in this commit as synthetic5 is the Random Permutation traffic
    • Misbah Mubarak's avatar
      Adding support for nearest neighbor comm. · 56007723
      Misbah Mubarak authored