| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 619.1 | Maybe 500 bridge | LACV01::TOMCZAK | | Tue Jun 23 1992 14:33 | 13 |
| Rolf,
You might pass that question to Tom Casey CSS::CASEY if he still is the
engineer.
A guess is that the "500" single port may work--since it has full
through put. On the microwave links the bridges really are for "trash"
filtering of the digitized RF (major use).
Last week I had a client ask the same question. My polite reply was why
worry about 6K$ in a 100K$+ dollar network.
Stan
|
| 619.2 | Thanks! | VAXRIO::ROLF | Vaporware Design Specialist | Tue Jun 23 1992 16:28 | 11 |
| Hi Stan!
Agree with the $6K / $100K statement, HOWEVER, we are dealing with the
brazilian Navy, and they use public funds... and they want answers to all
the IFs, WHYs and BUTs...
I always wondered what exactly the bridge in the MW box was good for.
Thanks for the pointer!
Rolf
|
| 619.3 | 3 port 600 bridge | LACV01::TOMCZAK | | Thu Jun 25 1992 13:56 | 9 |
| Rolf,
One last point. I understand that the 3-bridge module in the 600 unit
will, at times, with large traffic loads, cause the through-put to be
less than the single port bridge in the 500. Perhaps, someone out there
would know if this is accurate.
73's
Stan
|
| 619.4 | uh-uh | LEVERS::S_JACOBS | Live Free and Prosper | Thu Jun 25 1992 14:26 | 1 |
| No, this is wrong.
|
| 619.5 | clarification | LEVERS::S_JACOBS | Live Free and Prosper | Fri Jun 26 1992 10:57 | 15 |
| Paul sent me mail asking for clarification on this. If he needs
clarification, then it's safe to say others may as well.
The DB6XX bridges have a higher forwarding/filtering rate than the
DB5XX bridges, and the same translation rate. The DB6XX bridges do not
have enough forwarding/filtering to handle absolute worst case traffic
on all 3 ethernet ports. (It will gracefully degrade, not crash)
I thought the question was whether a DB6XX with just one ethernet port
connected will give the same performance as a DB5XX. The answer is
yes, but please make sure there are loopback connectors on the unused
ports. (avoids periodic packet delay/loss during self test of the
"broken" ports)
Steve
|
| 619.6 | any empirical work ? | MSBCS::KALKUNTE | Ram Kalkunte 293-5139 | Fri Jun 26 1992 20:52 | 6 |
| Has anyone measured throughput numbers with worst case traffic for the
DB5xx and DB6xx bridges ? This is the only way to answer this performance
question convincingly because worst case resource thrashing with the 2
bridges are different.
Ram
|
| 619.7 | DB600 performance paper indirect pointer | LEVERS::S_JACOBS | Live Free and Prosper | Wed Jul 01 1992 16:23 | 3 |
| re -.1
Please see note 286.7
|