Dell gig switches and autonegotiation nightmare. Am I alone here?

SVT4ME

Limp Gawd
Joined
Feb 2, 2006
Messages
221
I am now convinced it isn't me. I picked up some side work and have been deploying a new network for a client since last week. I purchased a Dell Powerconnect 48 port managed gig switch. Should be plug an play right? Well the servers have no issues since they have gig NICs. They talk to each other well, and file transfers between them blaze. What I AM seeing, however, is that any client with a 100 baseT NIC is running S L O W. I'm talking 2 minutes to copy 5 megs across the network. Now, I ran into this exact same issue about 6 months ago when I installed a Dell managed gig switch in another location. I can't consider this a coincidence. In both cases, putting the workstations back on the old (original) 100 baseT switch gets their speed back. I know most of you will say there is no point in having 20 100 baseT workstations on a Gig Switch, but I like the manageability, and I know the Dell has a ton more bandwidth on the backplane than their current Netgear 100 base switch.

I tried forcing the ports on the gig switch to 100, and the NIC on the workstation to 100 Full as well, and still no dice. What the hell am I missing? It should autonegotiate and just work damnit.

Anyone else seen this?
 
I am now convinced it isn't me. I picked up some side work and have been deploying a new network for a client since last week. I purchased a Dell Powerconnect 48 port managed gig switch. Should be plug an play right? Well the servers have no issues since they have gig NICs. They talk to each other well, and file transfers between them blaze. What I AM seeing, however, is that any client with a 100 baseT NIC is running S L O W. I'm talking 2 minutes to copy 5 megs across the network. Now, I ran into this exact same issue about 6 months ago when I installed a Dell managed gig switch in another location. I can't consider this a coincidence. In both cases, putting the workstations back on the old (original) 100 baseT switch gets their speed back. I know most of you will say there is no point in having 20 100 baseT workstations on a Gig Switch, but I like the manageability, and I know the Dell has a ton more bandwidth on the backplane than their current Netgear 100 base switch.

I tried forcing the ports on the gig switch to 100, and the NIC on the workstation to 100 Full as well, and still no dice. What the hell am I missing? It should autonegotiate and just work damnit.

Anyone else seen this?

I have seen symptoms similar to what you are describing and every time it has been due to poor wiring. Either the ends weren't wired to the A or B standard or the wiring had deteriorated so badly that it couldn't carry the signal. Doesn't even have to be a gig switch on the other end - new equipment just seems more touchy about the wiring (even 10/100).

(And yes, I looked a little goofy unwinding 100 feet of new 5e through the office to test, but it did prove the wiring was the issue.)

 
I have a 24 port 2724 with no issues, The only other thing I have seen other than "Party2go9820" wiring issue is a driver issue. Are all the workstations runnign the same NIC?
 
I have the identical problem, only the other way around. Currently, we have 2 Powerconnect 3448 switches (10/100 managed) serving our workstations. Most of the workstations are Latitude D600, 610 and 620's, all of which have built in gigabit network cards, however, many of the docking stations used are the older units which only support 10/100. I've tried about every setting I can think of, and the problem remains.

The only solution I've found is to move the patch to one of our gigabit switches. I currently have a couple new 48 port gigabit switches on their way from Dell to replaced the 3448's.
 
Well, if you can't get it fixed, I have 2 x brand new 3com 24 x 10/100 + 2 x 10/100/1000 managed switches for sale. I can make you a very reasonable offer on them if you'd like them. We bought the wrong models for my work, but by the time we realized our blunder the vendor wouldn't take them back.

Otherwise, I'd definetely test a fresh piece of known good cable. That sort of performance degradation is just ridiculous. Best of luck to you.
 
Back
Top