[9fans] 9P/Styx: streaming / contigious reading

[9fans] 9P/Styx: streaming / contigious reading

[Enrico Weigelt]

Hi folks,


did anyone already investigate how an streaming (w/o expicit read
requests) could be done via 9P ?

An option could be adding a new opcode telling the server that it
should send RRead's contigiously, within some given some range.
(will require an ordered transport, since RRead lacks an position
field, but quite trivial to implement).

What do you think about this ?


cu
--
----------------------------------------------------------------------
 Enrico Weigelt, metux IT service -- http://www.metux.de/

 phone:  +49 36207 519931  email: weigelt@metux.de
 mobile: +49 174 7066481   icq:   210169427         skype: nekrad666
----------------------------------------------------------------------
 Embedded-Linux / Portierung / Opensource-QM / Verteilte Systeme
----------------------------------------------------------------------

Re: [9fans] 9P/Styx: streaming / contigious reading

[hiro]

Why don't you use a protocol more suitable for high latencies?

On Mon, Aug 31, 2009 at 12:20 PM, Enrico Weigelt<weigelt@metux.de> wrote:
>
> Hi folks,
>
>
> did anyone already investigate how an streaming (w/o expicit read
> requests) could be done via 9P ?
>
> An option could be adding a new opcode telling the server that it
> should send RRead's contigiously, within some given some range.
> (will require an ordered transport, since RRead lacks an position
> field, but quite trivial to implement).
>
> What do you think about this ?
>
>
> cu
> --
> ----------------------------------------------------------------------
>  Enrico Weigelt, metux IT service -- http://www.metux.de/
>
>  phone:  +49 36207 519931  email: weigelt@metux.de
>  mobile: +49 174 7066481   icq:   210169427         skype: nekrad666
> ----------------------------------------------------------------------
>  Embedded-Linux / Portierung / Opensource-QM / Verteilte Systeme
> ----------------------------------------------------------------------
>
>

Re: [9fans] 9P/Styx: streaming / contigious reading

[erik quanstrom]

On Mon Aug 31 09:00:33 EDT 2009, 23hiro@googlemail.com wrote:
> Why don't you use a protocol more suitable for high latencies?
>

i think the problem rather is the tradition of having one
outstanding message per fid.  as far as i can tell, 9p doesn't
have this restriction.  we just use it that way.

- erik

Re: [9fans] 9P/Styx: streaming / contigious reading

[yaroslav]

> > > did anyone already investigate how an streaming (w/o expicit read
> > > requests) could be done via 9P ?
> >
> > Why don't you use a protocol more suitable for high latencies?
>
> i think the problem rather is the tradition of having one
> outstanding message per fid.  as far as i can tell, 9p doesn't
> have this restriction.  we just use it that way.

Maybe the way fcp(1) works is the answer?

-- Yaroslav

Re: [9fans] 9P/Styx: streaming / contigious reading

[Eric Van Hensbergen]

Was waiting for someone else to say it, but you should look at what
Octopus does with its operation continuation flags.

Another interesting twist on this is "lossy" streams -- but such a
thing may be best represented outside of 9P (or perhaps with a 9P
gateway).

    -eric

On Mon, Aug 31, 2009 at 9:20 AM, yaroslav<yarikos@gmail.com> wrote:
>> > > did anyone already investigate how an streaming (w/o expicit read
>> > > requests) could be done via 9P ?
>> >
>> > Why don't you use a protocol more suitable for high latencies?
>>
>> i think the problem rather is the tradition of having one
>> outstanding message per fid.  as far as i can tell, 9p doesn't
>> have this restriction.  we just use it that way.
>
> Maybe the way fcp(1) works is the answer?
>
> -- Yaroslav
>
>
>