Re: nfs inode busy message in 2.2.4

Tom Holroyd (tomh@nibh.go.jp)
Fri, 26 Mar 1999 16:42:57 +0900 (JST)


OK, I could reproduce it:

kernel: __nfs_fhget: inode 218771 busy, i_count=2, i_nlink=1
kernel: nfs_free_dentries: found //awajijo.jpg, d_count=0, hashed=1
kernel: nfs_dentry_delete: //awajijo.jpg: ino=218771, count=2, nlink=1

In this case, I copied the file to the NFS server (SunOS 5.4), and then
read it back a few times so it was buffered on the client (Linux 2.2.4).
Then I did some normal stuff for a few minutes, and then I deleted it from
the server side. A while later, I tried copying that same file back to
the server, though giving it a different name. At that moment, I got the
message above.

I never saw this before 2.2.4. That NFS server has an uptime of 52 days,
and doesn't do much else besides a couple of rc5's (dual CPU). The client
is an Alpha (PC164), 128M, no modules, Tulip ethernet. Oh, I did add the
arch/alpha/kernel/time.c patch to fix the negative gettimeofday deltas --
it works fine here, but I doubt it's relevant.

The /etc/fstab entry for the NFS mount reads:
bhmeg:/home/bhmeg/tomh /home/bhmeg/tomh nfs timeo=14,intr,rsize=8192,wsize=8192 0 0
and it's been mounted since I rebooted yesterday.

Dr. Tom Holroyd
I would dance and be merry,
Life would be a ding-a-derry,
If I only had a brain.
-- The Scarecrow

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/