We have a couple of Scientific Linux 6.1 NFS servers.  I looked at 
/proc/net/rpc/nfsd and was surprised to see on both of them that 
the thread histogram is all zero:

# cat /proc/net/rpc/nfsd | grep th
th 8 0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000


There are clients using one of these servers continuously, and at times
heavily, so I'm puzzled by these statistics and am assuming this is a
bug or configuration problem rather than a realistic depiction of the
thread state. 

For comparison, I looked a couple of our SL 5.x NFS servers and see
normal looking thread lines as expected, not just a bunch of zeros. 

Any thoughts on what could be going on?  Known bug?  Configuration
problem?  I'd appreciate if any other SL 6 admin with an active NFS 
server could check /proc/net/rpc/nfsd and see if the thread count 
histogram is empty like this.


If you have an SL 6 NFS server with non-zero thread stats, could you 
send the kernel and nfs-utils versions for comparison to ours, given below?

# uname -r
2.6.32-131.21.1.el6.x86_64

# rpm -qa |grep nfs
nfs4-acl-tools-0.3.3-5.el6.x86_64
nfs-utils-lib-1.1.5-3.el6.x86_64
nfs-utils-1.2.3-15.el6.0.sl6.x86_64


One of the servers has only NFS v3 clients (16 NFS client nodes), while
the other has only a single NFS v4 client.

Here is /proc/net/rpc/nfsd for one of them (sure to get line-wrapped
mangled):

# cat /proc/net/rpc/nfsd
rc 19 70434483 726969598
fh 352 0 0 0 0
io 2092104517 3971637111
th 8 0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
ra 32 393921819 0 0 0 0 0 0 0 0 0 985424
net 797404498 0 797447453 155310
rpc 797455746 0 0 0 0
proc2 18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
proc3 22 74845 228480507 19923612 17460391 64581715 115066 394911350
39971881 5313458 37750 36353 0 849734 3589 4288363 8251 267872 1146073
22996 140662 0 18601904
proc4 2 0 0
proc4ops 59 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0


Thanks for your time,

-Wayne Betts
 STAR Computing Support at BNL

 Physics Dept., Bldg. 510
 PO Box 5000
 Upton, NY 11973

 [email protected]
 631-344-3285


Reply via email to