We''ve seen a couple of instances where login/cluster nodes seem completely unresponsive when in a directory of a Lustre filesystem. When we''re in other non-lustre directories that very same node seems completely responsive. We''ve also correlated those evnets to very high load on the OSS Nodes (>250). Is there a way to figure out who''s beating up the filesystem so that we can help them use Lustre in a better way? Thanks, -- -Jason ------------------------------------------------- // Jason J. Hill // // National Center for Computational Sciences // // Oak Ridge National Laboratory // // e-mail: hilljj at ornl.gov // -------------------------------------------------