T.R | Title | User | Personal Name | Date | Lines |
---|
2386.1 | | STRWRS::KOCH_P | It never hurts to ask... | Thu Jun 15 1995 11:15 | 3 |
|
If I remember my DEChub 900 training correctly, repeaters connected
across the backplane do NOT increment the repeater count.
|
2386.2 | I count 4 repeaters. | CGOS01::DMARLOWE | Wow! Reality, what a concept! | Thu Jun 15 1995 16:27 | 12 |
| Multiple 900 repeaters on the same IMB are counted as one repeater
count. However as soon as you link a 900 repeater from a IMB to the
thinwire, to pick up a 90C or a 90FA, you start the count.
Since you have done this in both hubs you now have 4 repeaters and 5
coax segments. The number of coax segments allowed is 3 but since 2 of
the coax segments are the 90 thinwire backplane, they can be counted a
zero length segments.
Bottom line is that you are OK for now.
dave
|
2386.3 | It's valid. But not wise... | MSDOA::REED | John Reed @CBO = Network Services | Mon Jun 19 1995 10:56 | 15 |
| I would count that SPECIFIC config as 4 repeaters, and FIVE segments.
because the 90FA does not attach to an IMB, so you don't get the
benefit of the 900-series single-repeater functionality.
The Thickwire as SEgment 1, the Hub internal THINWIRE backplane does
count as a tapped segment, (and you have one in each hub) that that's
THREE tapped segments. you also have a link segment going to each
workstation, for a total of five segments. Your network is valid, and
at it's maximum extents.
If you used all 900 series hardware (and perhaps a fiber backbone) you
would reduce your segment counts.
JR
JR
|
2386.4 | Could be counted as 3 segments. | CGOS01::DMARLOWE | Wow! Reality, what a concept! | Tue Jun 20 1995 14:38 | 19 |
| >> I would count that SPECIFIC config as 4 repeaters, and FIVE segments.
Yes, to be safe.
>> the Hub internal THINWIRE backplane does
>> count as a tapped segment, (and you have one in each hub) that that's
I'm assuming that the same rules that apply to the 90 hub, apply to the
900 hub thinwire. If the 90 hub is terminated, it can be counted as a
ZERO length segment and effectively ignored. However, as soon as you add
a piece of thinwire to the hub, it then must be counted. The same should
apply to the 900 chasis as the thinwire is permanently terminated. If
that is the case the config would be 4 repeaters and 3 segments (one
coax and 2 link segment).
If Ed or someone from engineering disagrees, let me know.
dave
|
2386.5 | Not Ed or Eng. | KEIKI::WHITE | MIN(2�,FWIW) | Tue Jun 20 1995 15:34 | 6 |
|
Dave, I asked this back in Topic 1766 and the answers in 1766.3 and
1766.5 seem to disagree that you can disregard the length of the
thinwire connection in the DH90 or the DH900.
Bill
|
2386.6 | See DEChub 90 Owners Manual Page 2-2. | MSDOA::REED | John Reed @CBO = Network Services | Tue Jun 20 1995 17:05 | 15 |
| A zero length segment? I have never heard this. the rules that I
use to configure a 90-series hub is:
The internal impedance loss of each connector of the backplane create a
virtual thinwire equivalance of 15 stations, and 65 meters of ThinWire.
That is why you cannot attach more than 125 meters outside the hub, or
55 meters directly (no "T" connectors) to another hub.
I always count the backplane of a 90 Hub as a ThinWire segment. You
get to it through ONE repeater, and leave it through the SECOND
repeater.
JR
|
2386.7 | | STRWRS::KOCH_P | It never hurts to ask... | Tue Jun 20 1995 19:27 | 6 |
|
You can be creative with paired DEChub 90s. I've sold solution with
systems in the middle of DEChub 90 pairs and they've always worked. I
come out each one with a cable and connect them in the middle with a
T-connector to connect up systems and standalone module remembering not
to exceed the 185 meter Thinwire. Always worked for me...
|
2386.8 | Zero segment length. | CGOS01::DMARLOWE | Wow! Reality, what a concept! | Wed Jun 21 1995 17:23 | 20 |
| re. .5
Bill,
Note 1766.3 mostly references repeater hop count within the hub. It does
talk about the actual segment count.
Again I agree with the repeater count. Ed Buzzell's explaination with
regards to zero segment length has nothing to do with how many taps the
the hub thinwire has on it, be it 1 or 8. It has to do with length. If
the thinwire is contained within the hub, ie. terminated, the segment
is so short, ie. ~17 inches, that it can be ignored. If you add coax to
the side of a 90 hub it is no longer just 17 inches. Usually its many feet
so no longer that nice short internal thinwire. While contained in the hub
the collision domain is short that its length can be ignored and according
to Ed the segment can effectively be ignored.
Anyways, I'll leave it to Ed to correct me if the philosophy has changed.
dave
|
2386.9 | I base MY designs on MAX possible segment lengths. | MSDOA::REED | John Reed @CBO = Network Services | Mon Jun 26 1995 13:40 | 14 |
| I wouldn't want to assume that the customer will never expand that "T"
connector at the side of his DEChub 90.
I feel that it is important to provide a customer with a configuration
that has room to grow, and is not "MAX'ed out" at install time. He
should be able to buy a repeater module and insert it in any hub (you
know they will try...), or attach ThinWire to the end of an existing
terminator without causing my design to stop working.
It depends on the goal of your sale. If you want to meet an immediate
connectiviy need, or meet his architecture needs for his MIS strategy.
|
2386.10 | | CGOS01::DMARLOWE | Lost in thought without a map. | Tue Jun 27 1995 16:23 | 4 |
| You are sooo right.
dave
|