Skip to Content.
Sympa Menu

ndt-users - Re: Working with NDT under vmxnet: why is it 10 Mb?

Subject: ndt-users list created

List archive

Re: Working with NDT under vmxnet: why is it 10 Mb?


Chronological Thread 
  • From: Maurice Volaski <>
  • To: Richard Carlson <>
  • Cc:
  • Subject: Re: Working with NDT under vmxnet: why is it 10 Mb?
  • Date: Wed, 3 Oct 2007 13:56:09 -0400

* Turn off TCP Segmentation Offload (TSO) using "ethtool -K eth0 tso off"
* Set the NIC to interrupt after every packet using "ethtool -C eth0 rx-frames 1"

I am assuming you mean running these commands inside the hosted OS, not the guest OS. The don't work on guest OS driver, but it's not clear to me to what degree the vmxnet driver on the guest OS controls the computer's NIC. On the hosted OS, the first one was already true; the second one had been set to 5.

These will force the kernel to handle all the traffic, which is what should happen for this type of tester.

You can also send me part of the web100srv.log file. The "spds[?]" lines show the link detection timings. It would be good to see what they look like. Finally you can tell the NDT server to save some TCPDUMP files by adding the '-t' flag to the web100srv process. This would allow you to look at the packet arrival times to see what libpcap was doing.


I'm sending all separately; the duplex mismatch is back. :-(
--

Maurice Volaski,

Computing Support, Rose F. Kennedy Center
Albert Einstein College of Medicine of Yeshiva University



Archive powered by MHonArc 2.6.16.

Top of Page