JIVE Wiki

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
==== Minutes of NEXPReS WP6 MS601 telecon on 2010-11-02 ====

The telecon was held on 2010-11-02 and started at 11:00 UTC.

** Participants: **

Simon Casey (OSO), Neal Jackson (UMAN), Gerben van Malensteijn (SURFnet), Ari Mujinen (Metsahovi), Paul Boven (JIVE)

** Agenda: **

  - Progress on MS601
  - Any Other Business
  - Scheduling for a bi-weekly telecon

** Minutes: **

Gerben gave a short overview of the demo that was performed by SURFnet
and others at the GLIF Workshop at Cern in October, where dynamic
circuits amongst several NRENs were created, tested and torn down again
under software control. This included use of the Fenius protocol on
OpenDRAC for controlling SURFnet equipment and a Force-10 switch. Fenius
is an interim protocol for talking to Domain Controllers, which will
eventually be supplanted by the upcoming NSI protocol.

SURFnet (Gerben and Hans) will be adding text to the MS601 draft and
asked for an account for the WIKI.

Paul brought up the issue of IP assignment for dynamic paths. As the
end-point of the lightpaths at JIVE will be the central JIVE e-VLBI
switch/router, we should be able to assign VLANs (with virtual IP
addresses) to the ports that connect to the dynamic lightpath equipment.
At this moment, OpenDRAC only supports management of Force-10 switches,
but the JIVE switch/router is an HP ProCurve 5412zl. We should
investigate whether ProCurve switches can also be added to the supported
base of OpenDRAC.

Gerben mentioned that the next generation of SURFnet (SURFnet 7) is
currently being planned, and it is forseen that the current SDH based
backbone will be replaced by Metro Ethernet in this version. As the
rollout of SURFnet 7 will not happen before Q4 2011, we should plan on
using the current SDH based system first, and will have to adapt to the
new network somewhere during the NEXPReS project. As SURFnet 7 will also
be controlled through OpenDRAC, that should not be a major change from
the NEXPReS perspective.

Neal said he sees a distinction between 'hard real-time' requirements as
for e-VLBI, and the possibly less stringent requirements for LTA data
transfer, where it might be possible to adapt to available bandwidth, or
change scheduling to match bandwidth availability. Unfortunately the
scheduling of this telecon prevented Hanno Holties (ASTRON) from
participating and commenting on LOFAR's requirements.
Gerben replied that OpenDRAC has a 'find a timeslot' feature where it
will find the earliest timeslot that can accommodate a particular
request, although at the moment that feature isn't working.

Finally, Paul Boven (JIVE) proposed that, in view of the difficulties of
scheduling these telecoms, that it might be helpful to have a biweekly
telecon scheduled on a fixed time and day of the week. The participants
of the telecon preferred Wednesday mornings for such a bi-weekly
telecon, but of course the other participants in the WP will be asked
for their preferences too.

** Actions: **

Create Wiki accounts for SURFnet (done already).