[javaboyuk]

Hi May I make some comments?
Yes you should be able to change the read/wite size when mounting, see the man pages (welI I can on Solaris)

Anyway the real worry for me, your mounting the NFS mount with SOFT option, generally
you should NEVER do this if your going to write to it, and really unlikely even if its read only
unless you have multiple source to switch bettween, with specialist code to detect that etc.

So always mount NFS with HARD "hard" as the option. Now that is based on quite old
knowledge, but I can't see that level of fundamental approach changing :-)

So please mount it hard, the timeouts will still happen, but you will not loose data
ie get the IO errors your seeing, especally dangerous post flush() when the code
does not bother to check the return you effectively get silent data corruption.

Enjoy.
JB

Code:

OpenELEC:~ # mount | grep 10.150.213.2
10.150.213.2:/volume1/openelec on /storage type nfs (rw,noatime,vers=3,rsize=131072,wsize=131072,namlen=255,soft,nolock,proto=tcp,port=2049,timeo=3,retrans=2,sec=sys,local_lock=all,addr=10.150.213.2)

The NFS server is connected to the same switch as the Pi.

The timeouts seem to occur randomly but I can usually reproduce them by transferring a lot of data to/from the NFS mount at the same time:

Code:

OpenELEC:~ # dd if=/dev/zero of=/storage/test.data bs=8M count=512
512+0 records in
512+0 records out
4294967296 bytes (4.0GB) copied, 420.484441 seconds, 9.7MB/s
OpenELEC:~ # dmesg | grep nfs
OpenELEC:~ # cp /storage/test.data /storage/test2.data
cp: write error: Input/output error
OpenELEC:~ # dmesg | grep nfs
[136858.199309] nfs: server 10.150.213.2 not responding, timed out
[136858.207228] nfs: server 10.150.213.2 not responding, timed out
[136858.231235] nfs: server 10.150.213.2 not responding, timed out
[136858.231657] nfs: server 10.150.213.2 not responding, timed out
[136858.231918] nfs: server 10.150.213.2 not responding, timed out
[136858.248969] nfs: server 10.150.213.2 not responding, timed out
[136858.261939] nfs: server 10.150.213.2 not responding, timed out

What I've found is that if I change the rsize and wsize to smaller values, I don't have any problems:

Code:

OpenELEC:~ # mkdir /var/media/test
OpenELEC:~ # mount -o rw,noatime,vers=3,rsize=32768,wsize=32768,soft,nolock,proto=tcp,timeo=3,retrans=2 10.150.213.2:/volume1/openelec /var/media/test
OpenELEC:~ # mount | grep 10.150.213.2
10.150.213.2:/volume1/openelec on /storage type nfs (rw,noatime,vers=3,rsize=131072,wsize=131072,namlen=255,soft,nolock,proto=tcp,port=2049,timeo=3,retrans=2,sec=sys,local_lock=all,addr=10.150.213.2)
10.150.213.2:/volume1/openelec on /var/media/test type nfs (rw,noatime,vers=3,rsize=32768,wsize=32768,namlen=255,soft,nolock,proto=tcp,port=2049,timeo=3,retrans=2,sec=sys,local_lock=all,addr=10.150.213.2)
OpenELEC:~ # dmesg -c >/dev/null
OpenELEC:~ # cp /var/media/test/test.data /var/media/test2.data
OpenELEC:~ # dmesg | grep nfs

Is there a way to change the default wsize and rsize for NFS /storage when it's mounted via cmdline.txt?
[/quote]