Hi all,
To catch everyone up on the situation at Oxford, we've also now
applied the 'Liverpool' tcp settings to our pool nodes, and our
testing shows things working well (see Alessandra's blog post for
the details:
http://gridpp-storage.blogspot.co.uk/2013/02/sonar-test-to-bnl.html
). However, I've just been looking at our plots on the ATLAS
SONAR/FT monitoring pages:
http://bourricot.cern.ch/dq2/ftsmon/channel_view/multi/UKI-SOUTHGRID-OX-HEP&UKI-SOUTHGRID-OX-HEP&UKI-SOUTHGRID-OX-HEP&UKI-SOUTHGRID-OX-HEP&UKI-SOUTHGRID-OX-HEP&UKI-SOUTHGRID-OX-HEP/BNL-OSG2&TAIWAN-LCG2&RAL-LCG2&NIKHEF-ELPROD&SARA-MATRIX&IN2P3-CC/2013-02-13/2013-02-20/12/
and we don't actually seem to be moving very many large files to
BNL (arguably, that's not surprising since it's been pretty
broken) so we don't have continuous results on the plot, as
opposed to (e.g.) Glasgow who have results across that entire
period:
http://bourricot.cern.ch/dq2/ftsmon/channel_view/multi/UKI-SCOTGRID-GLASGOW&UKI-SCOTGRID-GLASGOW&UKI-SCOTGRID-GLASGOW&UKI-SCOTGRID-GLASGOW&UKI-SCOTGRID-GLASGOW&UKI-SCOTGRID-GLASGOW/BNL-OSG2&TAIWAN-LCG2&RAL-LCG2&NIKHEF-ELPROD&SARA-MATRIX&IN2P3-CC/2013-02-13/2013-02-20/12/
Do I/we/ATLAS/anyone need to do anything special to ensure or
encourage transfers between us and BNL now that we think we've
fixed things?
Ewan
|