My question is :
***************************************************************************
I have a file server with about 20 workstations using NFS mount to the
file server. On the client side, the following messages appear in the
message file quite often. However, no complaints from the users about
system problem. Patches 100075-09 and 100173-09 (NFS related) have been
installed in the file server.
Jul 23 09:47:13 orange vmunix: NFS write error 70 on host fruits fh 41b 1
a0000 1c130 33e030f6 a0000 2 5ea3500d
My question is :
1. What is the error regarding ? Will it cause any problems ?
2. Any other patches to install in the server and clients ?
3. Any way to eliminate this problem ?
Thanks!
******************************************************************************
Thanks for all those who responded to my questions. I have also
discovered that I am not the only one who is always getting this type of
errors. Here are the replies which I got.
code 70 crops up if a person is remotely logged on to the server in one
window and developing a program while testing it over NFS in another
window. (ie, a quick compile, then a quick test, ... )
" Also, error code 70 can occur when someone removes a file that a process
is actively writing from a NFS client machine."
NFS write error 70 is from a stale file handle. A file handle becomes
stale when the inodes to which it refers is freed or re-used. Example:
myclient% mkdir foobar; cd foobar
myserver% rmdir foobar
myclient% touch foo
touch: cannot create foo: Stale NFS file handle
Look up the error codes in /usr/include/sys/errno.h.
Two common NFS error codes are 13 - "permission denied" and 70 -
"stale file handle".
Error code 13 can occur from incorrect /etc/exports entry. Also,
it can occur because someone has changed the /etc/exports
entry to disallow the client after the client has already
been granted permission to perform this operation.
Error code 70 occurs when a fileserver gets rebooted and
the server NFS "file handle" for a particular filesystem
changes. The "file handle" can be changed under the following
circumstances:
A) Installing a new drive in place of an old.
B) Moving a filesystem from one devices to
another.
C) Performing a format, newfs, dump, and restore
cycle. Even if to the same device.
Get rid of the error by unmounting and remounting the
filesystem in question.
Also, error code 70 can occur when someone removes a file
that a process is actively writing from a NFS client machine.
Under SunOS 4.1, run "showfh" to translate the NFS
"file handle" given in the error message into a Unix pathname.
Beware that showfhd does a "find" on the server to get the
filename. "man 2 intro" will give more general information
on what error codes could mean.
A patch will be needed to get "showfh" to work correctly. The
patch id is 100371 and this patch is required for 4.1, 4.1.1,
and 4.1.2. Sun claims that it's been integrated into 4.1.3.
~~~~~~~~~~~~~~~~~~~~~~~the end~~~~~~~~~~~~~~~~~~~~~~~~~
Thanks again!
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:09:06 CDT