News:

• Pull request allowing separation of local and remote endpoints merged.
  • So, we are ready to start using local gateways for writes to ECHO! Any objections to this?

Operational issues:

• Sandbox store problems at CERN last week, especially during the weekend
  • Resulted in many spikes of rescheduled jobs and reduction of the number of running jobs across the grid
  • Fixed now
• power outage in the ECHO SN rack on Monday evening
  • Echo became sad, resulting in a spike of failed WGProduction jobs
• Intermittent ETF test failures for ECHO
  • One of the gateways (ceph-svc17) had *.s3.echo.stfc.ac.uk in its PTR record for the IPv4 address, as a result some requests were redirected to this incorrect endpoint
    • Fixed
• Failed downloads from RAL to HLTFarm
  • No external connectivity on HLTFarm WNs, therefore successful downloads are impossible. It's a bug in DIRAC -- these downloads should not have been attempted in the first place.