[pnfs] [nfsv4] Question on file layout!

[dean hildebrand]

On 10/19/07, Robert Gordon <rbg at openrbg.com> wrote:
>
> On Oct 18, 2007, at 5:50 PM, Suchit Kaura wrote:
>
> > I have been reading the spec and on page 272:
> >
> > http://tools.ietf.org/html/draft-ietf-nfsv4-
> > minorversion1-13#section-14
> >
> >
> > The following are very confusing if not inconsistent, I would really
> > appreciate your taking a few moments to look through and explain the
> > same:
> >
> >
> >     The number of elements in nfla_stripe_indices is always equal
> > to the
> > stripe count.
> >
> >    4.  nfl_fh_list.  An array of data server filehandles for each list
> >        of data servers in each element of the nflda_multipath_ds_list
> >        array.  The number of elements in nfl_fh_list MUST be of three
> >        values:
> >
> > ...
> >
> >        *  The stripe count, i.e. the number of elements as
> >           nflda_multipath_ds_list.  When issuing an I/O to any data
> >           server in nfla_multipath_ds_list[X], the filehandle in
> >           nfl_fh_list[X] MUST be used.
> >
> >
> > In the following example:
> >
> >       nflda_multipath_ds_list<> = { A, B, C, D }, { E }, { F, G }
> > stipe count = 3 ?
> >
> >       nflda_stripe_indices<> = { 2, 0, 1, 0 }
> > stripe count = 4? The example seems to take this route so the file
> > handle list should have 4 elements but it has
> >
> >       nfl_fh_list = { 0x36, 0x87, 0x67 }.
> > 3 elements?
>
> Suchit; It is a know draft issue and (not speaking for) the editors
> know about it.
>
> the example, of course, should read something like this.
>
>       nflda_stripe_indices<> = { 2, 0, 1}

Hmmmm, so how do we implement a pattern {2, 0, 1, 0} that repeats?  It
seems to me that is what the example was trying to illustrate.  The
table on page 282 demonstrates how to use  {2, 0, 1, 0}.  So I believe
the example is correct in d14, or is at least trying to demonstrate
how to handle a pattern where stripe_units > #devices.

Dean


> I put together a couple of pretty pictures over here in an effort to
> help illustrate.
>
> http://blogs.sun.com/macrbg/
>
> Robert.
>
>
> _______________________________________________
> nfsv4 mailing list
> nfsv4 at ietf.org
> https://www1.ietf.org/mailman/listinfo/nfsv4
>