<br><br><div class="gmail_quote">On Fri, Aug 7, 2009 at 7:51 AM, Dan Littlejohn <span dir="ltr"><<a href="mailto:dan.littlejohn@gmail.com">dan.littlejohn@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br><br><div class="gmail_quote"><div><div></div><div class="h5">On Fri, Aug 7, 2009 at 3:28 AM, belcampo <span dir="ltr"><<a href="mailto:belcampo@zonnet.nl" target="_blank">belcampo@zonnet.nl</a>></span> wrote:<br>
</div></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><div><div></div><div class="h5">
<div>Jim Stichnoth wrote:<br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
On Thu, Aug 6, 2009 at 11:05 AM, Dan Littlejohn<<a href="mailto:dan.littlejohn@gmail.com" target="_blank">dan.littlejohn@gmail.com</a>> wrote:<br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
I have a gentoo diskless frontend that mounts using nfs and it seems like<br>
the throughput of nfs is slow. iperf show the network able to push 770 Mb/s<br>
and hdparm shows the sata disk on the server at 90 Mb/s, but doing a dd<br>
across the network only shows up as 10 Mb/s (is about 40 Mb/s when the same<br>
test is done on the server to a third machine).<br>
<br>
Odd thing is that the pxeboot mounts nfsver2. Not sure if this is a problem<br>
or not, but other mounts after boot mount as nfsver3. Really seems like it<br>
is a config problem with nfs somewhere, but I have not found anything that<br>
makes a difference. Doing rsize=8k,wsize=8k, noatime, async. Anyone have<br>
any tips to look for or is this 10 Mb/s really the throughput I should<br>
expect for a diskless frontend?<br>
</blockquote></blockquote></div>
Remove those rsize=8k,wsize=8k, the default of current kernels is rsize=1048576,wsize=1048576<br>
Client/server negotiate max possible blocksize.<br>
If server is defaulting to a 'not optimum' block-size you can override that with<br>
echo 1048576 > /proc/fs/nfsd/max_block_size<br>
in one of your initscripts</div></div><div><div></div><div><div><div></div><div class="h5"><br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br>
My diskless MythDora 10.21 frontend uses NFS. I timed a copy of a 4GB<br>
file from the NFS server into /dev/null and got about 41 MB/s (or 330<br>
Mb/s). This is over a gigabit wired network. The only special mount<br>
option to speak of is "nolock" presumably because the lockd is not<br>
part of the initrd image.<br>
<br>
Jim<br>
_______________________________________________<br>
mythtv-users mailing list<br>
<a href="mailto:mythtv-users@mythtv.org" target="_blank">mythtv-users@mythtv.org</a><br>
<a href="http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users" target="_blank">http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users</a><br>
</blockquote>
<br>
_______________________________________________<br>
mythtv-users mailing list<br>
<a href="mailto:mythtv-users@mythtv.org" target="_blank">mythtv-users@mythtv.org</a><br>
<a href="http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users" target="_blank">http://mythtv.org/cgi-bin/mailman/listinfo/mythtv</a><br><br></div></div>Well, I got a bit farther. Letting the block size auto negociated made the write faster (Thanks Belcampo), but the read is still slow. Is the /etc/init.d/nfsd script where you set the proc max_block_size echo? I am running pxelinux so I believe I have to hard set the rsize and wsize in the pxelinux.cfg boot file, but it seems to make no difference to the read speed for different values I tried.<br>
<br>Here is a bit more info below with letting everything auto negociate with defaults including pxelinux when the root file system is mounted. Any other suggestions to try to find the read bottleneck? <br><br>Dan<br><br>
one ~ # nfsstat -m<br>/ from /dev/root<br> Flags: rw,vers=3,rsize=4096,wsize=4096,namlen=255,hard,nointr,nolock,proto=tcp,<br>timeo=600,retrans=2,sec=sys,addr=192.168.0.5<br><br>/store/tv from 192.168.99.99:/store/tv<br>
Flags: rw,noatime,vers=3,rsize=524288,wsize=524288,namlen=255,hard,nointr,noloc<br>k,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=192.168.99.99,mountvers=3,mountp<br>roto=tcp,addr=192.168.99.99<br><br>one ~ # time dd if=/dev/zero of=/store/tv/test bs=16k count=10000<br>
10000+0 records in<br>10000+0 records out<br>163840000 bytes (164 MB) copied, 2.79913 s, 58.5 MB/s<br><br>real 0m3.048s<br>user 0m0.033s<br>sys 0m1.983s<br>one ~ # time dd if=/store/tv/test of=/dev/null bs=16<br>
10240000+0 records in<br>10240000+0 records out<br>163840000 bytes (164 MB) copied, 15.548 s, 10.5 MB/s<br><br>real 0m15.558s<br>user 0m3.153s<br>sys 0m12.313s</div></div></blockquote></div></blockquote><div><br>
Dang, I left the "k" off my read test. Looks like I am good to go now. Thanks for the help.<br><br>Dan<br><br> one pxelinux.cfg # time dd if=/store/video/test of=/dev/null bs=16k<br>154614+1 records in<br>154614+1 records out<br>
2533208064 bytes (2.5 GB) copied, 56.4591 s, 44.9 MB/s<br><br>real 0m56.509s<br>user 0m0.163s<br>sys 0m26.802s<br><br></div></div><br>