kernel hanging problem - nfs?...

Wesley T. Perdue wes at
Thu Dec 26 20:04:00 CST 2002


At 11:34 AM 12/27/2002 +1000, jason andrade wrote:
>On Thu, 26 Dec 2002, Wesley T. Perdue wrote:
>> >are you using rsize=1024,wsize=1024 ?
>> No, 8192.  Here is a client fstab entry for the server:
>> nfsserver:/home  /home  nfs rsize=8192,wsize=8192,timeo=14,intr,hard,retry=1000000
>then please tell me you are using gigabit ethernet in all your clients and
>servers and you have switches that are known to support jumbo frames which
>is actually turned on.
>otherwise, you are stuffed if you try and use 8K nfs packets on linux under
>UDP.  i won't go into a huge amount of details as that'd make me want to
>bang my head against the wall some more to relive the pain.

Nope; just switched 100.  I had read somewhere that 8192 was ideal on a LAN.  So 1024 is ideal for switched 100, based on the 1500-byte Ethernet packet and the 1024-byte disk block, correct?

>> >nfsserver:/share /localmount nfs ro,rsize=1024,nolock,hard,bg,intr,retrans=8,nfsvers=3,timeo=10 0 0
>> >
>> >i use something like the above (add wsize and remove ro, since you'll want write access)
>> Thanks for the advice.  I'll look into the unique options you specify.
>best of luck.  also, you will want to remove "nolock".  i don't use any nfs locking because i only do read only nfs exports.

Got it.  Thanks again!

- Wes

More information about the Linux-PowerEdge mailing list