Attending: Sam, Winnie, Jens (chair+mins), Wenlong, Teng, Dan, Vip


Bristol not good.  Which is not good.  Like 700 processes, not high load, but they increasingly jam up.
Transfers fail, SAM tests.  Processes seem to not be able to write stuff.  Not high I/O.

It does start new processes but they don't do anything, or apparently don't have time to do anything.
Normally a filesystem issue, or bottleneck.  There is only one GridFTP node.  Maybe run fsprobe on it
(check with T1).  Something strange started ~15 Nov but gets more frequent.  Deploy another GridFTP server?

IRIS Workshop - lots of talks of status of digital assets.  How to run workflows in DiRAC.


https://indico.ph.qmul.ac.uk/indico/conferenceOtherViews.py?view=standard&confId=568

Need Rucio SE - Rob's setting up one, T1 should eventually provide a shared one.
We should try it out in January.

ops mtg yesterday - Alessandra pointed out we should investigate the efficiency of xroot proxy caches.
Mark should sign up - currently he has installed half of the required monitoring stuff.
Also need Teng's pages - still not visible due to network issue.
Cambridge - wil also have monitoring on it.

Sheffield - will also be running xroot proxy cache.


Looking back - pleasure - looking forward.  And Merry and Happy.