IGS bandwith OK

Thread View: ott.online

2 messages

[T] Back to Threads [V] Tree View [G] Group Articles

2 total messages Started by Ralph Doncaster Sat, 21 Jun 1997 00:00

#343

Author: Ralph Doncaster
Date: Sat, 21 Jun 1997 00:00

69 lines
3250 bytes

Hi.

I posted earlier in the week regarding apparent bandwith problems for
CA*Net traffic which I felt would also affect IGS (who gets their
feed from ONet).  Frank was nice enough to set up an account for me
so I could do some tests.

So far I've only had time to do one test.  It would appear that
either the CA*Net -> MCI link has been upgraded since the problems
I've seen, or traffic patters make it less used during off business
peak times (mon-fri 8am-6pm).  I'll post more test results as I get
them.  The interesting datapoint is the Wave service showing the
Wave->Istar bottleneck.  Both Magma and my pc on the Wave have to
go through IStar's MCI link, yet Magma was much better.  4.5K/s is
still a bit better than what you'd get on a modem, so I'm not about
to give it up (yet).
Lastly, note that one test is by no means statistically conclusive
of anything.

Sat June 21 16:35

Retrieving ftp.netscape.com/pub/smart/SM10R2.EXE.  All xfers started
at the same time (within 3 seconds of each other).

From doncaster.on.ca:
ftp>  mget SM*
local: SM10R2.EXE remote: SM10R2.EXE
200 PORT command successful.
150 Opening BINARY mode data connection for SM10R2.EXE (1409278 bytes).
226 Transfer complete.
1409278 bytes received in 15.9 secs (86 Kbytes/sec)
[Ralph's note: this is the kind of bandwith I like to see.  If only
sprint
 would learn how to keep it consistently this way...]

from host.ott.igs.net:
ftp> mget SM*
local: SM10R2.EXE remote: SM10R2.EXE
200 PORT command successful.
150 Opening BINARY mode data connection for SM10R2.EXE (1409278 bytes).
226 Transfer complete.
1409278 bytes received in 49 seconds (28 Kbytes/s)
[Ralph's note: note quite as fast as Frank's test, but adequate.  Even
though it would be more hops to use their UUNet link, my gut feeling
is that it would be faster if it were routed that way ]

from mag1.magma.ca:
ftp> mget SM*
local: SM10R2.EXE remote: SM10R2.EXE
200 PORT command successful.
150 Opening BINARY mode data connection for SM10R2.EXE (1409278 bytes).
226 Transfer complete.
1409278 bytes received in 66 seconds (21 Kbytes/s)
[Ralph's note: I guess it's clear Magma was really filling up it's
IStar T1 and needed the extra bandwith of their UUNet link]

from pc on Wave:
ftp> mget SM*
200 PORT command successful.
150 Opening BINARY mode data connection for SM10R2.EXE (1409278 bytes).
226 Transfer complete.
1409278 bytes received in 316.98 seconds (4.45 Kbytes/sec)
[Ralph's note: another one that would be better to go through UUNet;
Wave has both but seems to route most through IStar]

--
Ralph Doncaster, Doncaster Consulting Inc   email: ralph@doncaster.on.ca
2816 Richmond Rd. Ottawa, ON  K2B 6S5       ph:(613)829-6597

Re: IGS bandwith OK

#344

Author: Ralph Doncaster
Date: Sun, 22 Jun 1997 00:00

33 lines
1781 bytes

Ralph Doncaster wrote:
> from host.ott.igs.net:
> ftp> mget SM*
> local: SM10R2.EXE remote: SM10R2.EXE
> 200 PORT command successful.
> 150 Opening BINARY mode data connection for SM10R2.EXE (1409278
> bytes).
> 226 Transfer complete.
> 1409278 bytes received in 49 seconds (28 Kbytes/s)
> [Ralph's note: note quite as fast as Frank's test, but adequate.  Even
> though it would be more hops to use their UUNet link, my gut feeling
> is that it would be faster if it were routed that way ]

I guess I didn't proofread very well.  I meant to say that about
Magma.  IGS only feeds through ONet and doesn't have a UUnet link.
Magma is the one (below) with IStar and UUnet.

> 
> from mag1.magma.ca:
> ftp> mget SM*
> local: SM10R2.EXE remote: SM10R2.EXE
> 200 PORT command successful.
> 150 Opening BINARY mode data connection for SM10R2.EXE (1409278
> bytes).
> 226 Transfer complete.
> 1409278 bytes received in 66 seconds (21 Kbytes/s)
> [Ralph's note: I guess it's clear Magma was really filling up it's
> IStar T1 and needed the extra bandwith of their UUNet link]

--
Ralph Doncaster, Doncaster Consulting Inc   email: ralph@doncaster.on.ca
2816 Richmond Rd. Ottawa, ON  K2B 6S5       ph:(613)829-6597

Thread Navigation

This is a paginated view of messages in the thread with full content displayed inline.

Messages are displayed in chronological order, with the original post highlighted in green.

Use pagination controls to navigate through all messages in large threads.

Back to All Threads