Help.!! I Have a Stale NFS handle Problem.!! ;-{
 

Dear Reader,

I Have a Stale NFS handle Problem.!!
I never ran across a problem like this before.

My name is Michael E. and I Have been using Linux Slackware Since about 1995
I run Linux on all my Computers even My Laptop. I am using Slackware 12.2 Now.
Well Anyway, A while back I was a way from home and I wrote a letter to my wife on my linux laptop and sent it to her E-Mail. But I kept the Original Letter on my laptop I wrote it using Kword in KDE and Stored it as R-Molay.odt later I clean all my date file off my laptop and put them on a USB stick And forgot all about them. A few Days later I needed to save some stuff to my USB Stick but did not have the room So I Started removing old Files and then I get an Error that says: Can not remove R-Molay.odt Stale file handle So I do ls -al on the directory and get this

.
..
-????????? ? ? ? ? ?R-Molat.odt

I tried everything I could think of to get rid of it.. Dose anyone have any Ideas.?

Oh yes the Strange thing is that the file has No Data in it anymore. It is a Zero File. And it is on a USB Stick with a Linux ext2 Partition. I was Not part of a Network or an NFS system that I know of. It was just a letter I type in KDE using Kword on my Linux Laptop running Slackware 12.2 Dose any of this make since.? So How can a file that I created with data in it and saved to a linux partition on a USB Stick, windup with a Stale NFS handle and No Date in side of it.?? And the Big Question is How do I get rid of it off my USB Stick.!!

If anyone has any Ideas on how to get rid of a Stale NFS handle File,
Please let me know..

Thank you so much for your time.

Michael E. - Hard Core Power Linux User


<E077>

It's just how the error message goes. It means that particular entry is no longer correct. It can mean a program read the directory entry for a file, but the file was deleted before the program could read the file itself. It can also mean that you deleted the file, but something went wrong updating the entry for it (removed the stick early, disconnect due to loose connection, power outage, etc.). It may or may not be the sign of a bad sector that hit at just the wrong place. It's up to you whether to run fsck to correct the problem. You may want to run forensic tools to find or recover that file before using fsck, maybe not. If you didn't want that file, tinker with e2fsck, maybe read the man page for e2fsck to see if you want to try the safer options like -n and -p first.

This a moment to ask, "Do you have a backup?" Because I don't know for sure what caused the first bad entry, I don't know what a full fsck run will do, either. It might correct that one entry and be done, it might send one full directory to /lost+found. Read the e2fsck man page first, and if you agree with my assesment of "-n", try `e2fsck -n drive_to_check` after that.

Usually they will go away if you dismount the filesystem from the client, then remount on the client.

These files serve as a marker for files that had been open when deleted. The server needs them only briefly, but the client uses them with file handles. I have caused these by deleting a file on one client while another client was reading them.

Normally these pop into existence, and disappear with out anyone noticing.

As a worst case, you can use the "unlink" system call to remove it from the directory entry. Very likely there is no inode associated with it.