From: blockedofcourse@foo.invalid   
      
   I rescue lots of kit from the tip.  Presently, I am using 8 spindle   
   servers to replace (reimplement) my SAN appliances (the appliances   
   are closed source; file servers can be OPEN source!).   
      
   I was in the process of duplicating a spindle:   
      # cd /0; tar cpf - * | (cd /4; tar xpf - )   
   and belatedly decided to make some changes to the structure of the   
   source filesystem.  So, killed off that job and "rm -r /4/*".   
      
   This always mucks with the free block count which is annoying;   
   it's nice to df(1) and see identical block counts after a copy so:   
      # umount /4; newfs -U -j /dev/da4p1; mount /4   
   makes the disk look pristine.   
      
   OK, restart the copy:   
      # cd /0; tar cpf - * | (cd /4; tar xpf - )   
   And, lets see what sort of throughput we're getting:   
      # iostat -c5 da0   
   WTF?  *nothing*???  Literally *0* MB/s!   
      # iostat -c5 da4   
   Hmmm... 150MB/s.  That explains why the disk activity indicator for   
   drive 4 is blinking like crazy!  And, disk 0 is solid.   
      
   Open another session and I can see space being consumed on drive 4.   
   And, the actual contents from drive 0 appearing as time passes!   
      
   After some 3 minutes, the activity indicator for disk 0 starts   
   blinking...   
      
   Ah!  Cached files from the prior "copy".  Some 28GB of cached   
   files!  Well, it's good to see that the machine is making use   
   of that memory (96GB).  I wonder how much is set aside for the disk   
   cache vs. network and other kernel buffers, etc.   
      
   Quite a disturbing experience (did I type the command incorrectly?)   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)