UNSOLVED
Dell 3248 Locking up
I have a stack of 6 Dell 3248 switches. I have had three of them replaced under warranty for the same problem. One day I start to not be able to get SNMP messages from the switch. The next day I lose a port or two from the switch, then another, then another. I have the 2.1.0.0 code on the switches, just upgraded two weeks ago. Another switch is dying right now as I am typing this. I cannot ping, telnet, or console the thing. I have to unplug it from the wall for a bit and then plug it back in to work again.
I don't want to have to reboot this switch the rest of my life to ensure that it works. Does ANYONE know any kind of workaround for this issue?
Thank you so much!
James
Foxbat46552
8 Posts
217
0
Posted May 27th, 2004 17:00
James,
I'm afraid that you're in the same boat as us. The Management Engine on our PC 3248s have been subject to "locking up" since day one. As you can tell by the last Firmware date of July 2003, this is a Mature Product that Dell has no intention of maintaining, even with the reports that the PC 3248 continues to go "missing".
Other than applying V2.1.0.0, it was suggested that you disable IGMP MultiCast support, and enter the following into your "startup-config":
mcast mac-addr filter 01-00-0c-cc-cc-cd vlan 1
mcast mac-addr filter 01-00-0c-cc-cc-cc vlan 1
Of course, you would want to do this for each VLAN. Since we have done this, our incidence of PC 3248s becoming unmanageable has dropped to one or twice a year. Still unacceptable, as our Ciscos have been solid as neutronium. We are beginning the process of replacing our nine PC 3248s with comparable 48-port Cisco switches.
--Roland W. Fox
Thanks Foxbat.
I changed the layout so that now we only have three Dells in any given stack going to their own ethernet port. This hasn't helped us at all. I suppose I can try to do the mcast filter but that is painfully unnecessary. This experience has just taught me not to buy Dell networking products again. I have opened several cases with Dell support and they always give me the 'we have NEVER seen this before..it must be the way you are designing your network'. Three switches in a stack should be no problem for a hub that my baby daughter designs, much less for a company the size of Dell.
Anyway, thanks for the info. Now I know I am not crazy!
James
cpcizin
7 Posts
217
0
Posted April 22nd, 2004 18:00
I had the same problem for a long time. I also have 6 Powerconnect 3248's. My problem came down to basically overloading the switch, although when I looked at the statistics in the web interface, nothing was ever over 1% utilization. In any event, If you have all 48 ports connected to just computers, not a problem, however I had many other hubs and switches connected tojust one of the powerconnect 3248's and this would constantly initate the same problem you are facing.
The way I am setup now, is I have the 6 powerconnect 3248's daisy chained together via the Gigabit copper ports. Any additional hubs and switches I have are distributed somewhat evenly across the entire network. Basically one to three hubs/switches connected to a the ports in the Powerconnect in addition to the normal number of computers connected. This setup has been working out fine. So far my systems have been up for 158 days, 5 hours, 57 minutes, and 36.94 seconds without a problem! Basically, the person told me at tech support that the Powerconnect could easily handle the 48 ports, but when you plug in an additional 16 port switch, then an 8 port hub, and a 24 port hub, its like having the Powerconnect 3248 operate all 96 ports, which was overpowering the switch and cascading throughout the network and causing the problems.
Hope this helps!