[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 20011008: IDD latencies at PSU (cont.)

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: Re: 20011008: IDD latencies at PSU (cont.)
  • Date: Fri, 12 Oct 2001 04:48:38 -0600 (MDT)
Jeff,

We use iperf regularly for bandwidth testing.  I've started a server
running on;

  motherlode.ucar.edu (LDM server system in question)
  milton.unidata.ucar.edu (test system Anne suggested)
  
and agree that it may be appropriate to bump the window size to 64K
or more given the circumstances.

I'm not intimately familiar with the LDM code, but I'm relatively sure
that it doesn't do any sort of on-the-fly connection tuning.  On our
end, we turn up the system defaults on the (Solaris) LDM server. 

Are you working on the Web100 collaboration at PSC?

mike

> From address@hidden Thu Oct 11 21:48:31 2001
> Subject: Re: 20011008: IDD latencies at PSU (cont.) 
> 
> ... 
> You just need to get and compile iperf. You can get it from:
> http://dast.nlanr.net/Projects/Iperf/iperf-1.2.tar.gz
> 
> Then, you can run it on the machine with 'iperf -s -w 16K' for a server
> with a 16K TCP window. That window size is probably low for the rtt
> time between PSU and NCAR, but it should do for a start. I don't
> remember if LDM adjusts it's window or segment size.