Re: [9fans] Observation about sharing network/grid

Re: [9fans] Observation about sharing network/grid

[andrey mirtchovski]

> What a snide and insulting response.

Just like a school bully somebody has slapped in the face -- you jump on
everybody in sight looking for revenge...

> Keep thinking.

I will, but I refuse to do your thinking for you hereafter.

Re: [9fans] Observation about sharing network/grid

[boyd]

    Note to the world: I've finally started filtering out Choate.  Hooray!
    I get to join the rest of the masses in ignoring his inanities!

yeah, spamoff has a 'trash' function now.  i could just add it in or
change one line and bounce 'em.

Re: [9fans] Observation about sharing network/grid

[boyd]

    some of us have spent quite a bit of time thinking about it.

i'm with you captain ....

Re: [9fans] Observation about sharing network/grid

[Charles Forsyth]

>>NFS has been doing that for at least two yeas on Solaris, HP/UX, and AIX.
>>Might want to review the man pages. When I used to work for Tivoli we had

that would explain it.  our solaris system `pazzo' is older than that.
much had changed even then.  i see that `exporting' is out, and `sharing' is in.
quite.  very touchy-feely.  actually, that was one of the things about Sun's
Joyless announcement that worried me a bit:
something about him instilling something in
the hearts --and souls-- [emphasis mine] of those working for Sun.
what WAS going on there?  even my friends who worked for Microsoft
never spoke of anything like that.
(it was always about shares or share options with them.)

Re: [9fans] Observation about sharing network/grid

[Jim Choate]

On Sun, 5 Oct 2003, Charles Forsyth wrote:

> still, you're right: NFS could have done that, and perhaps the suenix
> re-implementations do.

NFS has been doing that for at least two yeas on Solaris, HP/UX, and AIX.
Might want to review the man pages. When I used to work for Tivoli we had
to make sure this was turned off because of some low level sys call
conflicts with Tivoli during installs from NFS images. Wasn't a big deal
since most sites didn't see it as very usefull.

In a public namespace/filesystem context however the importance changes
dramatically. If you don't allow transitivity in the mounts (and
lazy-update is another must have in the real world) then the source
servers get completely bogged down. Just imagine trying to export the
Gutenburg Project by forcing everyone to go to the project itself. By
using this technology a handfull of sites around the country could make
the primary connections and then export those out to their immediate
region. Then as users hit the second level server, only files that have
never been hit before (or have changed content even one bit) need be
re-propogated.

This whole 'a few sites share to another set of few sites' is the heart of
'small world networks'. It allows the network to scale without traffic
growing with it, provided the number of connections per node is not too
small or too large. Sort of like Goldilocks ;)


 -- --

God exists because mathematics is consistent, and the Devil exist because we
can't prove it.
                          Andre Weil, in H. Eves, Mathematical Circles Adieu

      ravage@ssz.com                            jchoate@open-forge.com
      www.ssz.com                               www.open-forge.com

Re: [9fans] Observation about sharing network/grid

[Charles Forsyth]

>>Just like NFS does. The transivity isn't in the control of the person

this touches an old nerve: NFS didn't (at least in Sun's implementation).
indeed, it limited exports to a physical file system, so that every machine
needed to know the mounting hierarchy of every other.  scales up well,
as you can well imagine!  anyhow, to get transitivity,
you need to work at it (patches possibly available on request, if i can find
them in my archive) and even then it doesn't work completely
(but it's much better than nothing), until you patch the clients...
one annoying thing is that it never became part of the traditional Jumbo Patch.

still, you're right: NFS could have done that, and perhaps the suenix
re-implementations do.

Re: [9fans] Observation about sharing network/grid

[Jim Choate]

On Sun, 5 Oct 2003, Dan Cross wrote:

> Uhh.  I import into a namespace.  I can export another namespace that
> doesn't include whatever I imported.  So import doesn't have to be
> `transitive,' as I said.

Dan, you're creating a strawman. Changing the point of discussion and then
claiming it's the same.

Your assertion as to what I said is simply wrong.

Further, what you do doesn't map to everyone else's needs.

So sad.


 -- --

God exists because mathematics is consistent, and the Devil exist because we
can't prove it.
                          Andre Weil, in H. Eves, Mathematical Circles Adieu

      ravage@ssz.com                            jchoate@open-forge.com
      www.ssz.com                               www.open-forge.com

Re: [9fans] Observation about sharing network/grid

[Dan Cross]

Jim Choate <ravage@einstein.ssz.com> writes:
> On Sun, 5 Oct 2003, Dan Cross wrote:
> > Umm, who ever said import was transitive?
>
> It's the export of the import, and Plan 9 has transivity built into it.
> Just like NFS does. The transivity isn't in the control of the person
> doing the export to the next person doing the export. That makes it
> transitive. Once the person can get the bits they can send them out
> without the source even being aware (eg proxy).

Uhh.  I import into a namespace.  I can export another namespace that
doesn't include whatever I imported.  So import doesn't have to be
`transitive,' as I said.  Your analogy with NFS doesn't make sense,
because in Unix-land (where I assume you're talking about), you have
a single, global namespace.  Not so in Plan 9.

> Further, without this sort of transivity the utility of grid computing
> (especially in a public sharing context) is pretty much nil. Among other
> things it causes way too much network traffic.

Uh, okay.

Note to the world: I've finally started filtering out Choate.  Hooray!
I get to join the rest of the masses in ignoring his inanities!

	- Dan C.

(Actually, I'm filtering him into another folder, so that I have something
to laugh at every now and again when I'm feeling bored.)

Re: [9fans] Observation about sharing network/grid

[Jim Choate]

On Sun, 5 Oct 2003, Dan Cross wrote:

> Umm, who ever said import was transitive?

It's the export of the import, and Plan 9 has transivity built into it.
Just like NFS does. The transivity isn't in the control of the person
doing the export to the next person doing the export. That makes it
transitive. Once the person can get the bits they can send them out
without the source even being aware (eg proxy).

Further, without this sort of transivity the utility of grid computing
(especially in a public sharing context) is pretty much nil. Among other
things it causes way too much network traffic.

 -- --

God exists because mathematics is consistent, and the Devil exist because we
can't prove it.
                          Andre Weil, in H. Eves, Mathematical Circles Adieu

      ravage@ssz.com                            jchoate@open-forge.com
      www.ssz.com                               www.open-forge.com

Re: [9fans] Observation about sharing network/grid

[Dan Cross]

Jim Choate <ravage@einstein.ssz.com> writes:
> It has been proposed to only import those grids one likes. Unfortunatley
> this won't work unless you're intending to demand that others limit their
> expansion choices to your choices. I suspect most people won't 'like' you
> very much on that point.
>
> Let's assume that we have three grids named A, B, & C.
>
> A doesn't 'like' C, but B does. A 'likes' B.
>
> A <-> B <-> C
>
> is equivalent through transivity to:
>
> A <-> C
>
> Back to the drawing board on that one....

Umm, who ever said import was transitive?

	- Dan C.

Re: [9fans] Observation about sharing network/grid

[Jim Choate]

On Sun, 5 Oct 2003, andrey mirtchovski wrote:

> That transitive closure would work in the non-Plan 9 world. Luckily what
> we're designing is better:
>
> In the 9grid we envision, if somebody is from C they are still considered
> as a C client for any/all authentication purpouses even though they may
> be coming from B, or from [AD-Z] for that matter. So filtering C
> /lib/ndb/auth-style is still going to work, with or without trusting B.

How would you know? Are you going to require that B authenticate all its
exports through A? That's going to require a fully connected network
between all nodes. Slow and tedious. Look into 'small networks'.

So you're going to prevent anyone from acting as a proxy? Won't work,
there are many sites that will require some degree of anonymity.

> some of us have spent quite a bit of time thinking about it.

What a snide and insulting response.

Keep thinking.


 -- --

God exists because mathematics is consistent, and the Devil exist because we
can't prove it.
                          Andre Weil, in H. Eves, Mathematical Circles Adieu

      ravage@ssz.com                            jchoate@open-forge.com
      www.ssz.com                               www.open-forge.com

Re: [9fans] Observation about sharing network/grid

[andrey mirtchovski]

That transitive closure would work in the non-Plan 9 world. Luckily what
we're designing is better:

In the 9grid we envision, if somebody is from C they are still considered
as a C client for any/all authentication purpouses even though they may
be coming from B, or from [AD-Z] for that matter. So filtering C
/lib/ndb/auth-style is still going to work, with or without trusting B.

some of us have spent quite a bit of time thinking about it.

andrey

ps: there's a 9grid list where this could be continued.

[9fans] Observation about sharing network/grid

[Jim Choate]

It has been proposed to only import those grids one likes. Unfortunatley
this won't work unless you're intending to demand that others limit their
expansion choices to your choices. I suspect most people won't 'like' you
very much on that point.

Let's assume that we have three grids named A, B, & C.

A doesn't 'like' C, but B does. A 'likes' B.

A <-> B <-> C

is equivalent through transivity to:

A <-> C

Back to the drawing board on that one....


 -- --

God exists because mathematics is consistent, and the Devil exist because we
can't prove it.
                          Andre Weil, in H. Eves, Mathematical Circles Adieu

      ravage@ssz.com                            jchoate@open-forge.com
      www.ssz.com                               www.open-forge.com