On our most heavilly loaded Samba server (running 1.9.18p10 under Solaris 2.5.1) we have got occaisional meltdowns with load average rocketing to 100+. The evidence indicates lots of smbd processes running. Syslogs indicate lots of: Nov 12 12:48:35 zeus.brunel.ac.uk smbd[21572]: Failed to set socket option SO_KE EPALIVE Nov 12 12:48:35 zeus.brunel.ac.uk smbd[21572]: Failed to set socket option TCP_N ODELAY Nov 12 12:48:41 zeus.brunel.ac.uk smbd[21573]: Failed to set socket option SO_KE EPALIVE Nov 12 12:48:41 zeus.brunel.ac.uk smbd[21573]: Failed to set socket option TCP_N ODELAY Also on the most recent incidence (happening once or twice a week), it was reportedhed that there were a lot of smbd connections bound to socket *.139 in the idle state. A glance at the source code indicates such failures ignored. Has anyone else experienced this? What more should I do to diagnose? Thanks, -- ----------------------------------------------------------------------------- | Peter Polkinghorne, Computer Centre, Brunel University, Uxbridge, UB8 3PH,| | Peter.Polkinghorne@brunel.ac.uk +44 1895 274000 x2561 UK | -----------------------------------------------------------------------------