perfSONAR User Q&A and Other Discussion

[Brian Candler <>]

On 20/09/2015 18:33, Brian Candler wrote:
> Like I say, I need to set this up again to test. It is possible that 
> it generated "send_only" for bwctl and "receive_only" for owamp (or 
> vice versa), but I think that would be a strange thing for the mesh 
> config to do.
I have confirmed this just using maddash inside a VM.

Given the following mesh config:

------------
description Example Mesh Config

<administrator>
  name       John Smith
  email      

</administrator>

<organization>
    description     Acme Examples, Inc

    <site>
        <host>
              description Acme East
              address pop1.sample.example
              address pop1.sample.example
        </host>
    </site>

     <site>
        <host>
              description Acme West
              address pop2.sample.example
              address pop2.sample.example
        </host>
    </site>
</organization>

<test_spec bwctl_test>
  type              perfsonarbuoy/bwctl  # Perform a bwctl test (i.e. 
achievable bandwidth)
  tool              bwctl/iperf3         # Use 'iperf' to do the 
bandwidh test
  protocol          tcp                  # Run a TCP bandwidth test
  interval          14400                # Run the test every 6 hours
  duration          20                   # Perform a 20 second test
  force_bidirectional 0                  # do bidirectional test
  random_start_percentage 10             # randomize start time
  omit_interval     5                    # ignore first few seconds of test
</test_spec>

<test_spec owamp_test>
  type              perfsonarbuoy/owamp  # Perform a constant 
low-bandwidth OWAMP test
  packet_interval   0.1                  # Send 10 packets every second 
(i.e. pause 0.1 seconds between each packet)
  loss_threshold    10                   # Wait no more than 10 seconds 
for a response
  session_count     10800                # Refresh the test every half 
hour (once every 18000 packets)
  sample_count      600                  # Send results back every 60 
seconds (once every 600 packets)
  packet_padding    0                    # The size of the packets (not 
including the IP/UDP headers)
  bucket_width      0.0001               # The granularity of the 
measurements
  force_bidirectional 0                  # do bidirectional test
</test_spec>


<group acme_bwctl_group>
    type       mesh

    member     pop1.sample.example
    member     pop2.sample.example
</group>

<group acme_owamp_group>
    type       mesh

    member     pop1.sample.example
    member     pop2.sample.example
</group>

<test>
  description       Example Throughput Testing
  group             acme_bwctl_group
  test_spec         bwctl_test
</test>

<test>
  description       Example OWAMP Testing
  group             acme_owamp_group
  test_spec         owamp_test
</test>
------------

After converting it to JSON, publishing it, and then using 
generate_configuration (with the machine configured so that its own 
hostname is "pop1.sample.example") I get the following in 
/opt/perfsonar_ps/regular_testing/etc/regular_testing.conf

------------
<test>
    <parameters>
        tool   iperf3
        omit_interval   5
        type   bwctl
        duration   20
        send_only   1
    </parameters>
    target   pop2.sample.example
    local_address   pop1.sample.example
    added_by_mesh   1
    description   Example Throughput Testing
    <schedule>
        type   regular_intervals
        interval   14400
    </schedule>
</test>
<test>
    <parameters>
        type   powstream
        send_only   1
    </parameters>
    target   pop2.sample.example
    local_address   pop1.sample.example
    added_by_mesh   1
    description   Example OWAMP Testing
    <schedule>
        type   streaming
    </schedule>
</test>
------------

So this confirms that it's using "send_only 1" for both bwctl and owamp. 
And therefore I would expect both the bwctl and owamp lines to be shown 
in the same direction (i.e. both solid or both dotted), but what I 
actually saw was that they were in opposite directions.

To test this properly of course I'd need to set up two complete nodes 
and configure them from a mesh, but maybe someone else has an existing 
mesh they can test quickly.

Or it might be quicker just to set "send_only 1" onto a couple of 
existing tests, and see how the graphs build. Unfortunately the GUI in 
3.4.2 doesn't let you set this flag (or even see it), but you can edit 
regular_testing.conf by hand.

Regards,

Brian.