Versions Compared

Old Version 12

changes.mady.by.user Behnam Montazeri

Saved on

compared with

New Version 13

changes.mady.by.user John Ousterhout

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

Voodoo Constants

  •  The round-trip time
  • The total number of available priorities
  • How should we divide the prios among the folowing packet types
    • Control packets (ie. grants)
    • Unscheduled packets (priority cut off among different message sizes)
    • Scheduled packets
    • Low priority redundant preemptive scheduled packets to avoid bubbles

Homa Paper ToDos

  1. Simulation comparison points ordered by importance

    1. self comparison when Homa features are removed

    2. pHost

    3. pFabric

    4. PIAS

    5. DCTCP

  2. Lower level measurement and analysisComplete the analysis of simulations with multiple senders and receivers:

    1. wasted bandwidth at receiver

    2. cumulative time average of priority usages

    3. time series of outstanding messages

  3. Algorithmic shortcomings

    1. we need not to use priorities for preempting large messages

    2. how should we pick parameters?

    3. need a module that is able to measure message size distribution

    4. how many priorities do we use and how many messages do we keep granted?if sender stops sending, should receiver "deactivate" message so it doesn't consume a priority level for redundancy?

    5. possibly receive multiple messages on the same priority level

    6. choosing parameters:

      1. dividing priorities between scheduled and unscheduled
      2. picking cutoffs for unscheduled priorities
      3. choosing redundancy factor for scheduled messages

    7. measuring actual workload, using it to pick parameters and change dynamically