[mythtv-users] NFS settings

[Mark Wormgoor]

Hi,

I am still waiting on my Myth machine, but am using NFS for homedirs and
some others already. Suggestions:
1. use rsize and wsize of 32768 for Myth, since most files are large (if
not huge) and you will mostly be doing sequential reads/writes.
2. if at all possible, use tcp and not udp; but your nfs server needs to
support that.  My options in fstab:
bg,hard,intr,tcp,nfsvers=3,rsize=32768,wsize=32768

Kind regards,

Mark

On Fri, 2004-10-08 at 22:48, Michael J. Lynch wrote:
> Don't know.  I know I use what is recommended and it works perfectly.
> 
> Joby Walker wrote:
> 
> > Yeah, saw that (I should have mentioned that), but there isn't any 
> > explanation of the settings.  I was wondering if people used different 
> > settings to effect.  21.9 recommends "soft" mounting, while the NFS 
> > documentation recommends "hard,intr".
> >
> > I assume a rsize and wsize of 8192 is because it is the largest block 
> > size without incurring packet fragmentation.
> >
> > jbw
> >
> > Michael J. Lynch wrote:
> >
> >> See the following
> >>
> >> http://www.mythtv.org/docs/mythtv-HOWTO-21.html#ss21.9
> >>
> >> Joby Walker wrote:
> >>
> >>> I am in the process of setting up an nfs server for my myth files, 
> >>> and I am curious about what people have done to optimize nfs 
> >>> performance.  For example what rsize and wsize.
> >>>
> >>> Thanks.
> >>>
> >>> jbw
> >>> _______________________________________________
> >>> mythtv-users mailing list
> >>> mythtv-users at mythtv.org
> >>> http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users
> >>>
> >>
> > _______________________________________________
> > mythtv-users mailing list
> > mythtv-users at mythtv.org
> > http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users
> >
-- 
***************************************************************
* |\    /|      |  /|  /       Mark Wormgoor                  *
* | \  / |      | / | /        mailto:mark at wormgoor.com       *
* |  \/  |ark   |/  |/ormgoor  http://www.wormgoor.com/mark/  *
***************************************************************