Attending: Jens, Luke, John B, Marcus, Rob C, Brian, Winnie, Govind, Steve

Only operational issues to discuss today:

* Luke's problem seems to be sorted via report to Andrea which prompted developers to provide a patch.
* Marcus' problem was discussed by email; it would seem the service does not support checksumming and
  affects, as described earlier, only when reading via xroot.

Problem seems to be with new ATLAS site mover which has bugs and does things in a different way from the
old one; xrdcp affected, not GridFTP [which probably doesn't checksum?]

Also the SL5, upgrading could be timed with hardware refresh but that may not be feasible for everyone;
draining a pool will take time, so the best option may be to upgrade the pool "live" - Brian is tracking
the few sites with resources on SL5.


John Bland: (23/11/2016 10:18:18)
don't need to drain a pool node to upgrade it
just be very careful ;0)