On Sat, 31 Dec 2005 22:52:29 GMT, CJT <(E-Mail Removed)> wrote:
>G wrote:
>> Hello,
>> Our workplace has just solved our networking
>> issues but I'm not quite sure as to the why we
>> had the problems, so I would like input/opinions
>> on the this post. I made this a brief as possible but
>> if you read it all it does get rather intriguing as you
>> go along.
>> In the beginning we had a linksys router and
>> a linksys hub and a cable modem. We had 7 nodes
>> total. Life was good.
>> Then we added two more nodes which was more
>> than our linksys router and switch had ports for and
>> also we changed ISP's from DSL to cable and they
>> brought out a router/cable modem combo, it was one
>> piece of equipment. They hooked up the coax cable
>> and left, telling us that they did not do network
>> maintenance. Well this is understandable so I started to
>> hook all of our peer to peer nodes that there was room
>> for into the new router/modem, and the rest into our old
>> linksys router Some nodes worked some did not. Our
>> troubles started.
>> The ISP router had LAN-1, LAN-2 etc thru LAN-5.
>> We thought maybe two routers were having trouble existing
>> together since there was no way to config the default address
>> of the ISP router/modem. So we bought a switch with enough
>> ports to handle all the nodes and we ran one cable back from
>> the switch to the ISP router/modem. We then could only see
>> 3 other random nodes from whichever node we were looking
>> from and none of the nodes could see out past the ISP router
>> to the internet, something that I still do not understand.
>> Being confused we called local "professionals". They examined
>> the issue and said we needed to buy (from them) Microsoft
>> Server software and a dedicated hardware server for a total cost
>> of $2500.oo including installation. We told them we needed a few
>> days to think about it. Something just did not seem right. So just
>> thinking of any scenario that we had not tried already we came up
>> with putting the old linksys router back into the scenario
>> "between" the strange acting ISP router/modem and our
>> switch. We ran all the nodes that would fit into the linksys
>> router leaving port one to receive the uplink cable from the
>> rest of the nodes in the switch, and also one port empty
>> next to the uplink port on the linksys from which we ran one
>> cable only to the LAN-1 of the ISP router/modem. We then
>> powered everything back up. After all the machines were
>> booted and ran about 2 min's or so, we started to test them and
>> to our pleasant surprise, everything worked on all nodes.
>> Internet, email, Lan, it all worked.
>> Life is once again good. But I'm still wondering what
>> exactly went down. I'm thinking the ISP router did not
>> have DHCP capability causing some of the nodes not
>> to work? Or maybe the ISP router does not broadcast
>> back to it's other ports and only takes them all onto the
>> cable line? I'm not sure what it was. I do know that with
>> the Linksys router (which does have DHCP capability)
>> in between our LAN and the ISP router/modem, things
>> work and without it they don't. I've read that today's
>> routers are "also" switches, were as originally they were
>> just routers. I know that switch it is more intelligent and
>> faster than a hub due to the fact that it can memorize the
>> destination addresses of all it's port nodes. Whereas a
>> hub simply broadcast all nodes. But a router simply
>> forwards all packets that would fall within a submask
>> range which (to me) would accomplish the same thing as
>> switch (maybe not as fast). In other words if the ISP
>> router/modem is only a router and not a also a switch it
>> would seem to me it still should have been able to get the
>> job done.
>> I know this is a long post but if any of you have
>> enough interest to decipher this I would greatly
>> appreciate your input in any way. I feel the more I
>> understand the better able I will be able to solve the
>> next crisis as our company grows.
>>
>>
>I wonder whether the perhaps the "gateways" were all set to the IP
>address of the Linksys, which might be different from the IP address
>of the cable router.
It sure sounds like a basic network addressing problem of some sort. I'm
wondering if it has to do with a FUBARed DHCP service, combined with not
power-cycling all of the equipment when making big topology changes.
With the modem/router wired only to a LAN port on the old Linksys router
(aside: one hopes that the OP isn't using the WAN port on the Linksys router)
which then feeds some nodes plus a switch feeding the rest of the nodes, I
believe this is the first configuration the OP has used that basically puts
all of the client nodes behind a *single DHCP server*.
It would be interesting to know:
- what ip address is each of the routers claiming? do they conflict?
- do both routers have DHCP enabled? (I bet yes).
- what ip address scopes are the DHCP servers using? do they overlap?
Given the equipment on hand, I'd make sure all systems are configured to use
DHCP, then power everything down, ditch the Linksys router, hook the switch up
directly to the modem/router, use the remaining ports on the switch and
modem/router to connect to the fleet, and then power everything back up. Life
should be good...
|