perfSONAR User Q&A and Other Discussion

[Shawn McKee <>]

Thanks Andy!

I have updated the three tests that use iperf3 by resetting the duration to 25 seconds.  Thanks for catching that.

Let's see how it goes now.

Shawn

On Wed, May 20, 2015 at 8:42 AM, Andrew Lake <> wrote:

Hi Winnie,

CCing Shawn McKee, since I think you will need him to update the mesh configuration file this host uses. I found a few things:

- Disabling the firewall got rid of the peer port issue. I think if you open 6001-6200 TCP and UDP you should be set to turn it back on.

- I was also able to reach your web page which allowed me to look at the mesh config file you are using at https://myosg.grid.iu.edu/pfmesh/mine/hostname/lcgnetmon02.phy.bris.ac.uk. The way this is configured, you are only running BWCTL throughput tests to hosts in the UK. You then have a very large mesh of traceroute tests you are running. That means for a whole bunch of those hosts listed on the graph page I wouldn’t expect any throughput data. My guess is that this is expected, but just pointing it out in case it’s not.

- For the throughput tests to the UK, most of those are failing and I think i see why (this is where we need Shawn). Your test configuration has the omit_interval set to 5 seconds and the duration set to 30. This is actually a 35 second test which is above the time limit of 30 seconds BWCTL allows by default. Thus the tests are getting rejected. This was actually a fix (I believe made around the April 26th date you noted prior) as previously BWCTL was only treating this as a 30 second test which was incorrect. The fix is to have Shawn change the “duration” in the meshconfig to 25, which will make it a true 30 second test within the BWCTL limits.

I think the summary is there are two problems 1) The firewall ports and 2) the test duration + omit_interval exceeding the BWCTL limit after a recent update. Hopefully resolving those will get things working.

Thanks,
Andy

> On May 20, 2015, at 6:55 AM, Winnie Lacesso <> wrote:
>
>
> Thank you Andy, this is very helpful!
>
> Earlier:
>> I can't seem to reach lcgnetmon02.phy.bris.ac.uk or run bwctl tests to it
>> from any host. My guess would be you have some firewall issues. The red
>
> For a time, the entire BW box firewall is shutdown. No firewall!
> (eeeek!)
>
> Can you try reaching lcgnetmon02.phy.bris.ac.uk or run bwctl tests to it
> again?
>
> While no firewall, still errors of the kind you named are being seen in
> /var/log/perfsonar/regular_testing.log:
>
> 2015/05/20 11:53:17 (3326) ERROR> MeasurementArchiveChild.pm:125
> perfSONAR_PS::RegularTesting::Master::MeasurementArchiveChild::__ANON__ -
> Problem handling test results: Problem storing results: Error writing
> metadata: Error running test from perfsonar02.datagrid.cea.fr to
> lcgnetmon02.phy.bris.ac.uk  with output bwctl: start_endpoint:
> 3628747485.110260
> bwctl: run_endpoint: receiver: 137.222.171.39
> bwctl: run_endpoint: sender: 192.54.207.251
> bwctl: Unable to initiate peer handshake with [137.222.171.39]:6033 - canceling
> bwctl: stop_endpoint: 3628747487.112685
>
> So how can the firewall be the problem, if it's off?
>
> Can you advise how further to debug what the problem is, if it's not
> firewall?
>
> Very Grateful!!
>
>