Skip to Content.
Sympa Menu

perfsonar-user - Re: [perfsonar-user] perfCube (Cubox-i4Pro with 16G microSD) packet loss problem

Subject: perfSONAR User Q&A and Other Discussion

List archive

Re: [perfsonar-user] perfCube (Cubox-i4Pro with 16G microSD) packet loss problem


Chronological Thread 
  • From: Mark Feit <>
  • To: Hyojoon Kim <>
  • Cc: "" <>
  • Subject: Re: [perfsonar-user] perfCube (Cubox-i4Pro with 16G microSD) packet loss problem
  • Date: Thu, 17 Sep 2015 13:07:00 +0000
  • Accept-language: en-US
  • Authentication-results: princeton.edu; dkim=none (message not signed) header.d=none;princeton.edu; dmarc=none action=none header.from=internet2.edu;

Hyojoon Kim writes:

My only concern was: would transmitting continuous logs through the interface that also does the continuous owamp influence the measurement results? I guess it depends on how much log is sent, including the size. Maybe the effect is negligible. I’m not sure myself, but yes, definitely also a good idea worth a try.

As we experiment more with low-capacity hardware with single network interfaces, we’re going to have to find ways to keep interfering traffic down to a minimum while tests are running.  Some of that may involve tweaking the logging system so it can be directed to buffer log data in memory while tests are running.  I suppose that if we wanted to be really clever about it, the logging system could reserve time on the interface to flush out what it accumulates.  I’d rather not go down that rabbit hole until there’s a good case for doing it.

—Mark




Archive powered by MHonArc 2.6.16.

Top of Page