On Apr 13, 2012, at 1:41 PM, Emmanuel Dreyfus wrote: > I run my perfuse/glusterfs tests on netbsd-6 branch, and I noticed that > I reclaims operations can lag for a long time. After running for a > while, perfused knows about 25217 nodes, most of them idle for hours. > Then kernel will start sending reclaims for old nodes when it wants to > lookup a new one.
VFS will cache up to kern.maxvnodes vnodes before it starts reclaiming vnodes to make space for new ones. > Is this an intended behavior? That consumes perfused and glusterfsd > memory for no benefit. Why is there no benefit of cached nodes? > Do I have a way to tell the kernel I would like a > more aggressive recycling policy? You could lower kern.maxvnodes then ... -- Juergen Hannken-Illjes - [email protected] - TU Braunschweig (Germany)
