NFS and nfsstat ...

NFS and nfsstat ...

Post by Thanh » Fri, 25 Feb 1994 13:51:27



During an NFS sesion, the following was captured:

On the client:
--------------
# nfsstat -c

Client rpc:
calls      badcalls   retrans    badxid     timeout    wait       newcred
188511     17029      0          42         17029      0          0          

Client nfs:
calls      badcalls   nclget     nclsleep
171542     0          171542     0          
null       getattr    setattr    root       lookup     readlink   read      
0  0%      9822  5%   0  0%      0  0%      95  0%     0  0%      280  0%    
wrcache    write      create     remove     rename     link       symlink    
0  0%      160371 93% 5  0%      4  0%      0  0%      0  0%      0  0%      
mkdir      rmdir      readdir    fsstat    
0  0%      0  0%      965  0%    0  0%      

The server shows no badcalls.

What does the above mean and how do I go about improve it ?

Thanks for help,

Thanh Ma,

--
--
Thanh Ma

 
 
 

NFS and nfsstat ...

Post by Dave Hi » Sat, 26 Feb 1994 18:07:03



>During an NFS sesion, the following was captured:

>On the client:
>--------------
># nfsstat -c

>Client rpc:
>calls      badcalls   retrans    badxid     timeout    wait       newcred
>188511     17029      0          42         17029      0          0          

>Client nfs:
>calls      badcalls   nclget     nclsleep
>171542     0          171542     0          
>null       getattr    setattr    root       lookup     readlink   read      
>0  0%      9822  5%   0  0%      0  0%      95  0%     0  0%      280  0%    
>wrcache    write      create     remove     rename     link       symlink    
>0  0%      160371 93% 5  0%      4  0%      0  0%      0  0%      0  0%      
>mkdir      rmdir      readdir    fsstat    
>0  0%      0  0%      965  0%    0  0%      

>The server shows no badcalls.

>What does the above mean and how do I go about improve it ?

All of your bad calls are actually time outs.  This just means that the
NFS server didn't respond to the request as quickly as the client
thought it should, so the client sent the request again.  Given that
writes are a time consuming request to service, and that your write
percentage is unusually high, my guess is that you are just overloading
your server.

You might try adjusting the NFS timeout parameter on your client's mount.
(This goes in /etc/fstab.  See mount(2) or mount(8) for details.)


Network Appliance Corporation                   (415) 428-5106

 
 
 

1. NFS Ver. 3 Statistics from a C program (like nfsstat does)

Hi,
        I am trying to get NFS Version 3 statistics from a C program under
HP-UX 11, AIX 4.3 and Solaris 2.6.  I would like to do like nfsstat does
to get at the stats.
        If someone can help, Please let me know how to get at these
statistics.

Thanks,

2. setup problem

3. Identifying NFS Timeout issue using NFSSTAT

4. 2.5.46: ide-cd cdrecord success report

5. nfsstat -m output and NFS mount options

6. RH 5.2 & sis6201 problem

7. Matrox Mystique ands X.

8. Has anyonr tried BOOTB before ? Boot from floppy B ?

9. nfsstat?

10. nfsstat for linux?

11. nfsstat for LINUX???

12. nfsstat equivalent for linux?

13. cltoomany (nfsstat)