Heya all, tickets incoming.
36 open UK tickets this week. Nothing too exciting other then a bunch of
tickets regarding unsupported glite software, most which have been
handled (and many would and have argued are non-issues, e.g. ClassicSE
entries). I've once again glossed over the various networking tickets,
but progress on those fronts is expectantly slow.
This ticket against the ROD complained about how old we "let" ticket
85973 get (Brunel's problem with the lcg utils timing out in EMI WNs).
The offending ticket is closed, so I think this ticket can be closed
too. Waiting for reply (15/10)
Sussex's high level of UNKNOWN status in September. John G has involved
NGI Ops and Kashif to ask why - I replied. In progress (15/10)
The September availability/reliability tickets for Glasgow & Durham.
Both sites have submitted a good explanation, i've asked if the powers
that be are satisfied. Waiting for reply (15/10)
Biomed are seeing the default job numbers in gstat, looks like dynamic
publish is broken. Assigned (12/10)
"Unsupported Glite Software" ticket. A plan is in the works for
upgrading. In progress (10/10)
An interesting one. LHCB jobs were failing at an Imperial CE due to the
node running out of inodes - not something I've seen before. Nothing
wrong with this ticket, just it caught my eye. In progress (11/10)
Oxford's torque server crashed, and lhcb noticed. Fixed quickly, so the
ticket can either be solved or lhcb can be asked if things are okay for
them now. In Progress (14/10)
Edinburgh has received a ticket about unsupported glite software at
their site. It's being handled though. In Progress (10/10)
Similar ticket for t'other side of Scotland. Sam makes some good points
about the "Classic SE" witchhunt, and asks for clarity on when the
actual deadlines are. In progress (11/10)
Liverpool's U.G.S. ticket. Steve gave it a cheeky In Progress, but no
other news (11/10).
LHCB are still seeing un-purgable jobs at Queen Mary, delivering another
list of zombie jobs. Has anyone else seen this? In progress (12/10)
That Sussex ticket :-) The nagios problems have been worked around,
Brian has been giving advice on Space Tokens. In Progress (11/10)
UCL's dpm problems were solved (touch wood) by increasing the VM
resources and correspondingly upping the mysql innodb_buffer_pool_size
setting. The lesson is don't skimp on your dpm headnode resources!
The Tier-1, Oxford, Cambridge, Bristol, Birmingham and Sheffield all got
U.G.S. tickets that they sorted promptly.
No other tickets of interest that I noticed, does anyone have any?