ndt-users list created

[Aaron Brown <>]

Hey Dave,

I've got some RPMs available if you'd like to test. There are x86-64 EL6 ones at http://ndb1.internet2.edu/~aaron/ndt-3.6.5.2-rc4.x86_64/ and i386 EL5 ones at http://ndb1.internet2.edu/~aaron/ndt-3.6.5.2-rc4.i386/ .

One thing to keep in mind is that web browsers and the java environment may cache the old version of the java Applet (for reasons that aren't remotely clear to me) so if you get the same results, try using a different computer that hasn't run the applet yet (or clear out caches).

Cheers,

Aaron

On Aug 21, 2013, at 11:34 AM, Dave Leibfritz <> wrote:

Does anyone know where the updated version is?

On 08/13/2013 08:24 PM, Richard Carlson wrote:

Hi Dave;

We found a bug in the code recently where the S2C byte counter was set
to an 32 bit int value.  At 10 Gbps rates this overflows with the
obvious results.

The latest version is fixed.  Somebody at Internet2 can tell you what
the updated version is.

Rich

On 08/13/2013 12:44 PM, David Leibfritz wrote:

When running NDT tests for 10 gigabit connections the client-to-server
results
are what we expect, 5+ gigabit/s, but the server-to-client is less than 1
gigabit/s.  Can you explain why or how we resolve this.  Running NDT
tests for
1 gigabit/s connections are consistent in both directions.  Below is an
example or our 10 gigabit NDT tests:

running 10s outbound test (client-to-server [C2S]) ..... 5709.95Mb/s
running 10s inbound test (server-to-client [S2C]) ......   898.46Mb/s

The slowest link in the end-to-end path is a 10 Gbps 10 Gigabit
Ethernet/OC-192 subnet

--

----------------------------------------------------------------------
David W. Leibfritz                          Phone:    (630) 252-9276
Accelerator Engineering Support Division    Email:
Argonne National Laboratory                 Fax:      (630) 252-6123
9700 South Cass Ave.                        Location: 401-C2244
Argonne, IL  60439-4844