Heikki Hannikainen <[EMAIL PROTECTED]> wrote:
> On Mon, 4 Jan 1999, Bob Nielsen wrote:
> > Joop, I see from the headers that your message took approximately one day
> > to arrive and went from vger.rutgers to listserv.funet.fi to my ISP (??).
> >
> > Why would something from the list be routed via funet?
>
> A lot of traffic from vger is routed via "exploder" hosts, like funet's
> mail server. vger's zmailer has been configured to punt a list of
> top-level domains to funet (one message transferred only, but with a HUGE
> recipient list), funet then makes the millions of connections to the
> recipient's mail servers. This is done just because vger isn't really up
> to the job, it's a tiny sparc classic box... mostly comparable with a 486.
These days the box is SparcStation10/50 -- 50 MHz SuperSPARC, and
128 kB RAM, but it still isn't up to speed to do about 100 000 smtp
sessions per day to varying locations, plus running Anon-CVS, plus..
Majordomo alone takes 10-20 seconds CPU per message to linux-kernel
list. Not a nice thing.
> This also distributes the cost of the bandwidth needed to transfer the
> list traffic (from whoever pays rutger univ's line, to the finnish
> government & taxpayers 8-). There are other exploder hosts for other
> TLD's too.
Numbers of different recipients per top-domain in all of VGER's lists are
listed here (top-10 in count), along with associated fanout relay:
....
633 ca vger.rutgers.edu
660 se ifi.uio.no
689 it nic.funet.fi
911 au samba.anu.edu.au
1176 org listserv.funet.fi
1226 uk ferret.lmh.ox.ac.uk
2634 de ferret.lmh.ox.ac.uk
2700 edu entropy.muc.muohio.edu
4475 net listserv.funet.fi
8119 com listserv.funet.fi
We *tried* to run ORG, COM, and NET at US located systems, but
when those relays failed, we moved the traffic back to FUNET...
> - Hessu
/Matti Aarnio <[EMAIL PROTECTED]>
<[EMAIL PROTECTED]>
