Monday February 23, 2009 at 12pm EST

EDIAL DETAILS:  Call: +1-734-615-7474 (PSTN CALL-OUT DOES NOT WORK)
or +1-866-411-0013 (toll free US/Canada Only)
Enter access code: 0196692

Attendees:

  • Mike Van Norman - UCLA
  • Carla Hunt - MCNC
  • Jason Zurawski - Internet2
  • Aaron Brown - Internet2
  • Joe Metzger - ESnet
  • Jeff Boote - Internet2
  • Prasad Calyam - OARnet
  • Andy Lake - Internet2

Agenda:

  • Roll Call
  • Current Activities (All)
  • Opportunities to get involved in the working group:

(1) Contributing to a mission statement for the working group
(2) Contributing to a "pretty good measurement practices" document
(3) Giving an update on current measurement activities in your world on a working group conference call
(4) Sharing your organization's experiences deploying Internet2 performance tools; whether those experiences are positive or negative
(5) Contributing ideas towards developing training targeted towards key users of measurement infrastructure. For example, we have discussed the possibility of creating a You Tube like video about measurement tools that NOC staff could watch when it is convenient for them.

  • Action Items (if applicable)

Notes:
Carla begins discussion talking about recent experience of deploying measurement infrastructure around MCNC. 10 severs at MCNC pops currently, running: ndt, owamp, bwctl. Looking to add more tools at some point. Early in the process still (happy to have the equipment out first).

MVN: What are you doing w/ regards to measurement conflicts (e.g. how to stop bwctl from effecting ndt tests, etc.)
CH: We will need to consider this in the future, we are not doing anything specific right now.
JB: We have talked about trying to integrate NDT into BWCTL scheduling to prevent this sort of problem - have only talked about this as a high level idea thus far, there are no plans beyond the talking stage.

Carla asks Prasad to give an update on activities at OARnet. Prasad talks a little about measurement expereiences and current infrastructure. Had used BWCTL at some point, using something now that is similar to PiPES (very simple tools ping, iperf, no NDT currently, scheduling infrastructure and analysis to get notifications). Have some measurement hosts on backbone (~6 right now, was as high as 20 some time ago). Mentions that his Ph.D. thesis was on conflict avoidance and adjusting forcasting tools such as NWS.

Carla asks MVN a similar question. Mike explains that there UCLA in particular uses OWAMP/BWCTL/NDT. Calren uses similar tools, CENIC mostly the same (not all currently - would need to check). Also admits documentation on what tools are where are out of date and he will need to update it.

Carla mentions an upcoming conference call dealing with TL1 monitoring for USLHCNET. Details will be shared when they are available.

Carla talks about some ways forward for the group:

  1. Creating a mission statement
  2. measurement best practices
  3. Getting others involved

There were no immediate offers of help on the call about these topics, Carla will solicit help via the mailing list and future meetings.

Jeff gave a brief update on Internet2 Activities:

  • Working on Cisco Telepresence verification tools (derivative of the tools available on the NPToolkit). This is still an early project - more info some other time.
  • Monitoring activities on the DCN pilot
  • Upcoming releases of tools:
    • NPToolkit
    • perfSONAR-PS
    • OWAMP
    • BWCTL

Call next month - details will be discussed on the mailing list.

  • No labels