Skip to Content.
Sympa Menu

perfsonar-user - Re: [perfsonar-user] Xeon-D adequate for low cost 10G PS under $2,000 USD? If not, recommend cheapest 10G HW.

Subject: perfSONAR User Q&A and Other Discussion

List archive

Re: [perfsonar-user] Xeon-D adequate for low cost 10G PS under $2,000 USD? If not, recommend cheapest 10G HW.


Chronological Thread 
  • From: Paul Gunn <>
  • To: Mark Feit <>
  • Cc: Minseok Jang <>, "" <>
  • Subject: Re: [perfsonar-user] Xeon-D adequate for low cost 10G PS under $2,000 USD? If not, recommend cheapest 10G HW.
  • Date: Wed, 1 May 2019 08:55:27 +1200

I totally agree with pscheduler's wisdom on this one. This was more of a bench test to see what the hardware would do.

Me: "I want to do something daft!"
pscheduler "Computer says no"

pg

Paul Gunn
Systems Administrator

Research & Education
Advanced Network New Zealand

M +64 21 521 884
E


On Wed, 1 May 2019 at 01:16, Mark Feit <> wrote:

Paul Gunn writes:

 

I tried to do a simultaneous 10G test between two 10G nodes, 22 ms apart, to show you. But pscheduler was too smart for me. First one, then the other.

 

pScheduler restricts that because running more than one of certain kinds of test at a time runs the risk of distorting the results for all of them.  Running simultaneous throughput in both directions may not be the bests thing to do because making return traffic swim upstream will affect how TCP behaves.

 

If there are tools that can do a bidirectional measurement, they can be integrated.

 

--Mark

 




Archive powered by MHonArc 2.6.19.

Top of Page