HP-UX 11i June 2002 Release Notes

New and Changed Disk and File Management Features
Network File System Support on TCP/IP
Chapter 11
198
JFS 3.3 uses new header files. As far as the JFS module is concerned, a well-behaved
application will not need to be recompiled. However, a kernel-intrusive application will
need to be recompiled with the new header files, and possibly with some corresponding
code changes. You should check with the application provider before upgrading.
Performance Issues
JFS 3.3 generally outperforms previous releases, as design changes have reduced the
number of bottlenecks resulting from globally shared locks. JFS includes tunables and
features to support improved performance in the OLTP, DSS, and technical computing
markets. With appropriate tuning, JFS 3.3 also outperforms HFS in all categories.
JFS 3.3 includes a new command, vxtunefs, for tuning a VxFS file system. See vxtunefs
(1M) and tunefstab (4). Also see the HP JFS 3.3 and HP OnLineJFS 3.3 VERITAS File
System 3.3 System Administrator's Guide for information on tuning a JFS file system.
NOTE The volcopy and labelit commands will be obsoleted in a future release. You should
use vxdump and vxrestore for backup and restore, or you can use an application-specific
utility. You can use dd to make a literal copy of the file system.
Network File System Support on TCP/IP
new at 11i
original release
With versions 2 and 3, Network File System (NFS) is now supported over the
connection-oriented protocol, TCP/IP, in addition to running over User Datagram
Protocol (UDP).
As a result of this new functionality, NFS is now supported over wide-area networks
(WANs). As long as TCP is supported on the WAN, then NFS is supported also. (TCP
transport increases dependability on WANs. Generally, packets are successfully
delivered more consistently because TCP provides congestion control and error recovery.)
The mount_nfs command now supports a proto= option on the command line where the
value for proto can be either UDP or TCP. (In the past, this option was ignored.) This
change allows administrators to specify which transport protocol they wish to use when
mounting a remote file system.
If the proto= option is not specified, then NFS, by default, will attempt a TCP
connection. If that fails, it will then try a UDP connection. Thus, by default, you will
begin using TCP instead of UDP for NFS traffic when you begin using the 11i version of
HP-UX. This should have little impact you. You do, however, have the option to specify
either UDP or TCP connections.
If you specify a proto= option, only the specified protocol will be attempted. If the server
does not support the specified protocol, the mount will fail.
The nfsd daemon now opens TCP transport endpoints to receive incoming TCP requests.
For TCP, nfsd is multi-threaded. For UDP, nfsd is still multi-processed.