Meeting of HEPiX IPv6 working group. Thursday 14th July 2011, 16:00-17:00 CEST (EVO video conference) Attendees: Thomas Finnern, David Foster, David Kelsey, Soumaya Lanouar, Mark Mitchell, Kars Ohrenberg, Andreas Pfeiffer, Francesco Prelz, Ronald van der Pol, Luuk Uljee, Ramiro Voicu. 1. Notes and issues from last meeting (22 June 2011). DavidK thanked Edoardo (not present) and Soumaya for providing some notes. These had been uploaded to today's agenda page. David apologised that he did not manage to add any notes of his own, although the slides worked on during the last meeting (and uploaded to the agenda page for the 22 June) present many of our decisions as to the work plan for the year ahead. In time this will be transferred to the new group wiki as a permanent reference. No issues were raised. 2. DavidK reported that he had given a short presentation on our working group to the WLCG Collaboration Workshop at DESY (11-13 July). See https://indico.desy.de/materialDisplay.py?contribId=43&sessionId=9&materialId= slides&confId=4019 3. Main discussion of the meeting was on the distributed IPv6 testbed. We agreed at the last meeting that sites involved would have at least one dual stack system available for this (running SL5 and with a trusted, i.e. not self-signed, X.509 Host Certificate), with routing to a public IPv6 network. It was confirmed that at least the following sites plan to be in the early testbed (31st August): CERN, DESY, INFN, Manchester, Glasgow, Caltech (USLHCNET machines at CERN), KIT, EPFL and SARA. INFN and the Caltech/CERN machines already exist and during the meeting it was confirmed that INFN could ping6 the Caltech machines. These machines were however not yet running SL5. Mark Mitchell reported that he had worked with Sabah at Manchester to set up a IPv6 tunnel between the two sites. This had stalled pending resolution of some configuration issues at Manchester. The issue between Manchester and Glasgow is that the Glasgow side requires a target ipv4 address for the ipv6 to ipv4 tunnel, to be configured at Manchester. Sabah is looking into this as his router presently is configured for IPv6 only. In the meantime, Glasgow University networking had responded positively to a request to route IPv6 packets for this testing. We discussed whether to have special DNS names for testbed machines and agreed that this was not necessary. We did however agree that having dns entries of form : -V4.domain (to reach IPv4 endpoint) and - v6.domain (to reach IPv6 endpoint) would be very useful for testing. We agreed that machine names would be recorded in the IPv6 wiki. Francesco Prelz will take care of setting this up if needed during next two weeks while DaveK is on holiday (see next section). We discussed whether it would be possible to have a single X.509 host certificate with multiple host names (to include the -V4 and -V6 variants). Francesco confirmed that he had looked into this and it was not possible (problems with Grid middleware). This is a topic for further investigation perhaps. 4. IPv6 Wiki The address of the new IPv6 wiki is: http://w3.hepix.org/ipv6-bis We need to request usernames and passwords for individuals to have write access. DaveK will request these and contact people. 5. Date of next meeting and AOB. At the June F2F meeting we said that we would plan to have another face to face meeting at CERN at the start of September. Two possible dates then were 1st Sep or 7th Sep. Today we agreed that 1st Sep does not work for some so we agreed *tentative* dates for this next face to face meeting: ** Tues/Wed 6th/7th September 2011 at CERN (subject to confirmation by Edoardo at CERN) ** (The plan would be to start at a time on Tuesday morning to allow most people to travel that morning and then to finish early enough on Wednesday afternoon to allow people to return home - exact times to be confirmed). We will try to confirm this meeting early in August. Given the holiday season we agreed that there is little point having another video conference before this face to face meeting. We did note that some were having problems with audio in the EVO conference so agreed to try a pure phone conference next time. Notes by David Kelsey 15th July 2011