[Beowulf] automount on high ports

Robert G. Brown rgb at phy.duke.edu
Wed Jul 2 08:46:51 PDT 2008

On Wed, 2 Jul 2008, Perry E. Metzger wrote:

> You don't switch to a different port number after the connection comes
> in, you stay on it. You can in theory talk to up to (nearly) 2^48
> different foreign host/port combos off of local port 25, because every
> remote host/remote port pair makes for a different 4-tuple.

Ah.  I should have known that.

>> Many daemons have a limit that can be set on the number of
>> simultaneous connections they can manage.
> That's a resource issue, not a TCP architecture issue per se. You
> might not have enough memory, CPU, etc. to handle more than a certain
> number of connections.
> By the way, you can now design daemons to handle tens of thousands of
> simultaneous connections with clean event driven design on a modern
> multiprocessor with plenty of memory. This is way off topic, though.

Not on a cluster list.  Networking in a very real sense IS the topic.
I've written forking daemons (which is why I should have known, or
remembered, about the four-tuple thing:-) because they are an essential
component of IPCs in a network-based cluster or cluster distributed
apps.  Even though PVM and MPI make it easy to write portable code (and
may well provide you with better performance than you can easily get on
your own) there may well be occasions for cluster software writers to
need to write their own networking, in band or out of band.

>> However, this is for TCP ports that maintain a persistent connection.
>> UDP ports are "connectionless" and hence somewhat different.
> I'm assuming they're doing NFS over TCP. If they're using UDP, things
> are somewhat different because of the existence of "connectionless"
> UDP. However, they *should* use TCP for performance. (I know people
> used to claim the opposite, but it turns out you really want TCP so
> you get proper congestion control.)

Yah.  To make UDP reliable, you have to load it down with most of the
stuff in TCP anyway; it isn't clear that it was ever a great choice.
IIRC PVM was originally built on UDP for similar reasons, but I think --
am not sure but think -- it is TCP today because it wasn't worth the
hassle.  I'm too lazy to crank up a PVM app to find out, though...;-)


Robert G. Brown                            Phone(cell): 1-919-280-8443
Duke University Physics Dept, Box 90305
Durham, N.C. 27708-0305
Web: http://www.phy.duke.edu/~rgb
Book of Lilith Website: http://www.phy.duke.edu/~rgb/Lilith/Lilith.php
Lulu Bookstore: http://stores.lulu.com/store.php?fAcctID=877977

More information about the Beowulf mailing list