Some SAM test SRM/tape endpoint failures on Monday/Tuesday. Castor team found there were a lot of requests in the system and it was having an impact on all VOs. Possibly caused by some tests being performed by LHCb, reading from Castor buffer, but there were also a lot of reads from NA62.
Job efficiency is going through a good period, this is due to a job mix including a very large proportion of GEN-SIM jobs which do not use significant I/O.
VO Liaison LHCb1m
(Science and Technology Facilities Council STFC (GB))