Circuit TCP-0-2, Adjacent node = 19.41 (SG1)
SG1 is the only node flapping, so I assume the problem is closer to SG1 than
STUPI?
$ multinet traceroute 69.21.253.158
traceroute to 69.21.253.158 (69.21.253.158), 30 hops max, 38 byte packets
1 192.108.200.210 (192.108.200.210) 10 ms 10 ms 10 ms
2 194.71.10.95 (194.71.10.95) 10 ms 10 ms 10 ms
3 kst5-core-1.tengige0-2-0-2.tele2.net (130.244.64.100) 10 ms 10 ms 10 ms
4 bck-core-1.bundle-ether4.tele2.net (130.244.39.141) 10 ms 20 ms 20 ms
5 gbg-core-2.bundle-ether2.tele2.net (130.244.39.52) 20 ms 20 ms 20 ms
6 gbg-core-2.bundle-ether5.tele2.net (130.244.49.166) 30 ms 20 ms 20 ms
7 ams-core-1.bundle-pos1.tele2.net (130.244.205.150) 30 ms 30 ms 40 ms
8 nyc9-core-1.pos8-0-0.tele2.net (130.244.218.214) 110 ms 110 ms 110 ms
9 nyc9-peer-1.xe-1-2-0-unit0.tele2.net (130.244.64.138) 110 ms 110 ms 110 m
s
10 198.32.118.31 (198.32.118.31) 100 ms 110 ms 110 ms
11 69.21.253.154 (69.21.253.154) 160 ms 140 ms 150 ms
12 * * *
No, I don't remember winprinters. On the other hand, I have not really
moved beyond RSX yet... :-)
Winprinters were a short-lived, very stupid idea for consumer-grade
printing from several different manufacturers. They did all the
rasterization and scaling within the device driver (which was, of
course, only available for Windows...this was in the 1990s) running on
the host computer. The printer itself contained little more than the
mechanical frame and the electrical drivers for the print head's jets,
and the stepper motors.
"Winmodems" were similarly-stupid garbage that did all the DSP on the
(Windows) host.
Thank heaven that drek was short-lived.
With todays network gear, you will not get most packets to your machine
anyway, the switches are already doing the filtering for you, so
promiscuous mode don't hurt much from that point either.
...and this is the only thing that makes me more comfortable with the
idea. The switch will just have more than one entry in its MAC table
for this port.
Right. Which is no different than if you cascade two switches. The
switch is perfectly happy, and your machine will only get the packets
you care about, and the random packet for a new machine, and the random
packets that goes to no machine (which should be pretty close to zero).
So my Ethernet frame filtering gets moved from my desktop system's
network interface, instead of "inward" to the IP stack, "outward" to the
Ethernet switch's MAC filter...where it's being done anyway. :)
But I STILL don't like it! B-) *snicker*
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
No, I don't remember winprinters. On the other hand, I have not really moved beyond RSX yet... :-)
With todays network gear, you will not get most packets to your machine
anyway, the switches are already doing the filtering for you, so
promiscuous mode don't hurt much from that point either.
...and this is the only thing that makes me more comfortable with the
idea. The switch will just have more than one entry in its MAC table
for this port.
Right. Which is no different than if you cascade two switches. The switch is perfectly happy, and your machine will only get the packets you care about, and the random packet for a new machine, and the random packets that goes to no machine (which should be pretty close to zero).
But I still don't like it. ;)
:-)
Johnny
