• Added 3 C6420 Nodes (288 cores) to be shared between UM Tier3 and Tier2.
• Site started draining (65% usage) on 6th May, reported to ADC, added a second gatekeeper, it started to ramp up after 24 hours.
• Had low transfer efficiency and job stage in error (27% job failure) over the weekend, file access time out. The solution was to restart all dCache services.
• Update dCache from 6.2.15 to 6.2.21 on Monday; Seems to have helped: greatly decreased transfer errors and increased job efficiency.
• Will also test adding 50% more memory to one of the MSU dcache pool nodes with the most pools and files as those still seem to cause more errors than they should.
• Update 3 squid  servers (UM site) to 4.15-1.1 to address recent security alert; MSU has updated one squid server to SL7, another one to be updated.