Skip to Content.
Sympa Menu

perfsonar-user - Re: [perfsonar-user] throughput test timeouts

Subject: perfSONAR User Q&A and Other Discussion

List archive

Re: [perfsonar-user] throughput test timeouts


Chronological Thread 
    < Chronological >      < Thread >
  • From: Mark Feit <>
  • To: Michael Thomas <>, "Garnizov, Ivan" <>, "" <>
  • Subject: Re: [perfsonar-user] throughput test timeouts
  • Date: Fri, 11 Mar 2022 14:46:15 +0000
  • Arc-authentication-results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=internet2.edu; dmarc=pass action=none header.from=internet2.edu; dkim=pass header.d=internet2.edu; arc=none
  • Arc-message-signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-AntiSpam-MessageData-ChunkCount:X-MS-Exchange-AntiSpam-MessageData-0:X-MS-Exchange-AntiSpam-MessageData-1; bh=8yLf2/mUOPW35YAMnyTDmUXVnB5tvQVZMzHnglYdAzA=; b=P9yxxg1A8d/1xmo3dVlKjSyp3bq7mzOzjzdU0OmFXgYKwuRRCkd31S3EMLf7XbkKonpq9H1FSrvjM2pzp/c8NAUK7vkZB9zD9Xxr9+vc58cBz9f+0jxEgGxBzCdbN+k+U2zPuaY9uOvJnIlir/0yfQBvVoRj7AfvXjYIh9h5mu6BXIH8inJmQ1+zrfDLNSFM3T+1T3pBWmPPf368QLjTVT3Al3ykbas5D3Po1qHmXdOthMyeqEAesvAgHG7hZIFVuJBlLQnPqeiAD3pCOUkLyV6OtC90nnwzfJhSK9wNP4L96zjGXBzyT+wm6NiydY77BpdymO332HfKygxFEHLa8g==
  • Arc-seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=ZcoexFYm8z2IjLNK5vzipNxmoETXaofJLYVyulCF25vW+E73jYEIqRk3+LtX2hg4PpuWmo4NtkZRSXnu/CBnCQ+EZXFFirHHYpPvg+xSbyCZ82mttNDu4gpd6tqBprZNk6I4rtpo2SJ+C1vQgDve0K2tviDoFwx1G3YvqY8g/WCzaVcHt7iRAruEoz9BD/EyG+WOmWFJ15dP+Zn0RSE97WAuiZDRy73XE3YY61A/V30SLscG5fm5a7XbTMdYg5flXr4/dxs863jyq+HWA0ObgSIVkXz74VWjYw3iSmIdy133+Y9Hu04RIuHhILKm5GMhQvGFEMQ8wqUk/os6vyFv6g==

Michael Thomas writes:

 

This is in fact the problematic behavior.  The iperf3 process does not
go away after each throughput test.    It does appear that the stray
iperf3 process gets killed when the next throughput test starts, but the
test will launch a new iperf3 process that stays around after the test
is complete.

We’re seeing this on some of the hosts run by the development partners and are looking at ways to mitigate it.

pScheduler’s plugin runs the iperf3 server in one-shot mode, which means it will exit after a successful test.  What I suspect may be happening is that when the server runs and no client connects, it isn’t being properly killed off when the plugin’s timeout elapses without the process having exited.  We have a library function that handles running programs for all of pScheduler; it may just need a tweak to be more aggressive after timeouts.

--Mark

 


Archive powered by MHonArc 2.6.24.

Top of Page