Skip to Content.
Sympa Menu

perfsonar-user - [perfsonar-user] Specifying reverse iperf3 throughput tests in a psconfig template

Subject: perfSONAR User Q&A and Other Discussion

List archive

[perfsonar-user] Specifying reverse iperf3 throughput tests in a psconfig template


Chronological Thread 
  • From: "Uhl, George D. (GSFC-423.0)[SGT INC]" <>
  • To: "" <>
  • Subject: [perfsonar-user] Specifying reverse iperf3 throughput tests in a psconfig template
  • Date: Wed, 8 May 2019 16:49:56 +0000
  • Dkim-filter: OpenDKIM Filter v2.11.0 ndjsvnpf101.ndc.nasa.gov 72926401CB58

Hi,

 

I’ve not been able to find a psconfig template variable in the throughput test type specification that allows me to request that iperf3 tests run with the “—reverse” option between a managed host and a non-agent host.  I know the flip variable is available for latencybg tests.  Can it be used in throughput tests with the same behavior?

 

I’m running into occasional cases where iperf3 tests sourced from the non-agent host are failing but they succeed if they’re run from the managed host when the reverse option is set.  Controlling bi-directional tests from a managed host to a non-agent host would work for me in these cases.

 

Thanks,

George Uhl

 

From: <> on behalf of "David J. Chaffin" <>
Reply-To: "David J. Chaffin" <>
Date: Wednesday, May 1, 2019 at 10:34 AM
To: Minseok Jang <>, Paul Gunn <>
Cc: "" <>
Subject: Re: [perfsonar-user] Xeon-D adequate for low cost 10G PS under $2,000 USD? If not, recommend cheapest 10G HW.

 

Related:

 

Though Paul's data show D15xx should be fine for 10Gb, has anyone tried Epyc 3xxx?  Price of Xeon-D, maybe 3-4x performance.  Maybe it could do 25/40/100 Gb perfsonar? 

Example:

 

--

David Chaffin

U. of Arkansas

 

 

 

 

 

 

 


From: <> on behalf of Paul Gunn <>
Sent: Monday, April 29, 2019 8:28 PM
To: Minseok Jang
Cc:
Subject: Re: [perfsonar-user] Xeon-D adequate for low cost 10G PS under $2,000 USD? If not, recommend cheapest 10G HW.

 

Hi Minseok,

 

I have ~25 perfSONAR testpoints installed around New Zealand based on Supermicro 5018D-FN8T systems, or that motherboard in a different case (DC data centres). Their X10SDV-TP8F motherboard is based around a Xeon processor D-1518. And they work fine for me.

 

We're running on Ubuntu 16.04, with only the testpoint packages installed on the machines.

 

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.

 

Running iperf3 from the command line on both hosts simultaneously, i get ~9.82 Gbits/s

 

pg.


Paul Gunn
Systems Administrator

Research & Education
Advanced Network New Zealand


M +64 21 521 884

E

 

 

On Tue, 9 Apr 2019 at 13:51, Minseok Jang <> wrote:

Hi, all.

 

I'm planning to deploy 10~30 PS nodes in KREONET until next year. Most sites are connected with 1G, but I want to deploy 10G PS for the future. A budget for 1 PS node is under $2000 USD.

 

I found that Xeon-D processor contains 10GbE controller in the processor[1]. Supermicro Xeon-D solution has IPMI remote console. It is important for remote management. Its price is around $1500 at Amazon[2]. Unfortunately, I found that Xeon D-1540 gives poor performance on this video[3]. It shows 2.6Gbps on Windows 10 in the iperf test.

 

Is there any perfSONAR based one Xeon D-15xx or Xeon D-21xx? Is it recommendable for 10G perfSONAR? If not, could you recommend low cost one?

 

Best Regards

Minseok Jang

 

 

--

Minseok Jang

 

KREONET Center / Supercomputing Division

Korea Institute of Science and Technology Information

 

Email:

Phone: +82-42-869-1292 / Mobile: +82-10-2630-7480

--
To unsubscribe from this list: https://lists.internet2.edu/sympa/signoff/perfsonar-user




Archive powered by MHonArc 2.6.19.

Top of Page