Monday 28 April 2008

Impact of a SPOF


Last week connectivity to the RAL Tier-1 was affected for many hours due to a large number of firewall connections being left open. The ops SAM tests quickly showed how the Tier-1 can act as a Single Point of Failure with the current setup. The evidence is in this gridpp grid status snapshot in the 24hrs column. Note that Dublin and ScotGrid sites were unaffected as they now run their own RB/WMS and top-level BDII.