Add a sysctl to specify the amount of UDP receive space NFS should

    reserve, in maximal NFS packets.  Originally only 2 packets worth of
    space was reserved.  The default is now 4, which appears to greatly
    improve performance for slow to mid-speed machines on gigabit networks.

    Add documentation and correct some prior documentation.

Problem Researched by: Andrew Gallatin <gallatin@cs.duke.edu>
Approved by: jkh

1 files changed, 8 insertions, 1 deletions

diff --git a/share/man/man5/rc.conf.5 b/share/man/man5/rc.conf.5
index 95aa2be..4ef54a0 100644
--- a/share/man/man5/rc.conf.5
+++ b/share/man/man5/rc.conf.5
@@ -491,10 +491,17 @@ It is commonly needed to run CFS without real NFS used.
 .Ar YES ,
 allow services like PCNFSD to make non-privileged mount
 requests.
-.It Ar nfs_reserved_port_only
+.It Ar nfs_privport
 (bool) If set to
 .Ar YES ,
 provide NFS services only on a secure port.
+.It Ar nfs_bufpackets
+(integer) If set to a number, indicates the number of packets worth of
+socket buffer space to reserve on an NFS client.  If set to
+.Ar DEFAULT ,
+the kernel default is used (typically 4).  Using a higher number may be
+useful on gigabit networks to improve performance.  The minimum value is
+2 and the maximum is 64.
 .It Ar rcp_lockd_enable
 (bool) If set to
 .Ar YES