Fix a flipped comparison in the memory pressure calculation
authorMichael Vrable <mvrable@cs.ucsd.edu>
Thu, 31 Mar 2011 22:06:14 +0000 (15:06 -0700)
committerMichael Vrable <mvrable@cs.ucsd.edu>
Thu, 31 Mar 2011 22:06:14 +0000 (15:06 -0700)
We were almost always saying there was too much dirty data that needed to
be flushed.  Now, also print out debugging messages when memory pressure is
detected stating what we have found, exactly.


No differences found