1. SUMMARY OF SIMILAR ERRORS
Just want to comment that I had similar error messages showing up in dmesg (see 3/1 SL post) 
Re: NFS client issue and kernel 2.6.18-348.1.1.el5

where the nfs client was SL 6.1 (64 bit) and the server was SL 5.9 (32 bit).

similar but not quite the same as
https://access.redhat.com/knowledge/solutions/306063

With the SL 5.9 buggy kernel 2.6.18-348.1.1.el5...
In my case, I found one missing directory (not on client, but on server).
All files in troublesome directories and 2 of 3 subdirectories listed 3 times each. They are listed properly, one time each, on NFS server file system.

Using an older kernel on the SL 5.9 server solved the problem 2.6.18-308.24.1.el5

boot of server (SL5.9) with broken kernel 2.6.18-348.1 showed RPC.idmap or idmapd (don't recall) with RED ERROR message. I don't remember seeing this when rebooting with older kernel that worked 2.6.18-308.24.1.el5.

2. I have yet to try a newer kernel on 5.9 to see if problem solved with either SL5.9 server update or SL6.1 client update. However seeing this exchange, I think I'll either not update kernels (by placing exclude=kernel* in yum.conf file) or keep them all around (installonly_limit = 0) and keep booting the older OK kernel. 

Bill Lutter