Skip to Content.
Sympa Menu

edugain-discuss - Re: [eduGAIN-discuss] eduGAIN Connectivity Check Service for individual SPs

edugain-discuss AT lists.geant.org

Subject: An open discussion list for topics related to the eduGAIN interfederation service.

List archive

Re: [eduGAIN-discuss] eduGAIN Connectivity Check Service for individual SPs


Chronological Thread 
  • From: Chris Phillips <Chris.Phillips AT canarie.ca>
  • To: "edugain-discuss AT lists.geant.org" <edugain-discuss AT lists.geant.org>
  • Subject: Re: [eduGAIN-discuss] eduGAIN Connectivity Check Service for individual SPs
  • Date: Fri, 3 Mar 2017 15:30:41 +0000
  • Accept-language: en-US

Hi Jim,

Yes, a good example! I think there's some interesting datapoints to be
drawn from that environment:


Some thoughts:

My recollection of it's use is point to point testing as opposed to full
on 'let's saturate everything because we'll see what breaks' testing.
Am I mistaken?
I'm poking at their dashboard:
https://fasterdata.es.net/performance-testing/perfsonar/perfsonar-dashboard
/

And see test suites that stay within geo-political boundaries (with
the
exception of APAN: http://ps2.jp.apan.net/maddash-webui/ )

Also this is more a qualitative test of quality of the linkage regardless
of what's on the other end of the wire
(e.g. I can reach a region or endpoint of the network as opposed to a
'Service of the network')

Additionally, a unique aspect of perfSONAR is that one needs to deploy a
dedicated VM or physical machine capable of doing adequate tests.
This means that the testing endpoints are single purpose as opposed to
multipurpose or carry other services.

I think that's an important distinction.



Other observations: the matrix model of the dashboards are n x m but both
numbers are 'small'.

I'm not seeing 'all known perfSONAR boxes to all other perfSONAR boxes'

https://www.perfsonar.net/use/locating-other-instances/


(opening http://stats.es.net/ServicesDirectory/ says 'this may take
several minutes' so I think they too have the problem)



A good case to look at..
C



On 2017-03-03, 10:09 AM, "Basney, Jim" <jbasney AT illinois.edu> wrote:

>On 3/3/17, 8:52 AM, Chris Phillips wrote:
>> Are there cases of other internet scale components that do this? I'm
>> skeptical that there are and they probably don't do it for a reason.
>
>I offer perfSONAR as a model:
>http://www.perfsonar.net/about/who-uses-perfsonar/
>
>Having perfSONAR probes operating globally has been extremely useful for
>detecting and diagnosing wide-area networking problems impacting large
>scale science projects.
>
>-Jim
>

Attachment: smime.p7s
Description: S/MIME cryptographic signature




Archive powered by MHonArc 2.6.19.

Top of Page