Printable View
I am having trouble with 24 port unmanaged fast ethernet swithes supporting static IP users locking up from time to time. This will often be shortly after a power loss, but the switch is protected by a UPS. This doesn't seem to happen with the DHCP switches. Does anyone have any idea what the problem is and how to fix it?
Do you have access to any sort of logs on this unit? Maybe that could give some insight into what's causing it. Or, perhaps the UPS unit is defective. My suspicion is that it's sending a small surge when switching from AC to battery power and causing the switch to lock up.
Switch ould also be defective. If it's under warranty, I'd have it replaced or try a different brand.
Thnaks for responding. I have 3 switches connected to the same UPS, but only the one dedicated to static IPs locks up. I've had the problem with Dell, Sysco and NetGear switches. I've got quite a collection now. After being unplugged for a day, they can be put back into service.Quote:
Originally Posted by dcomiskey
Changing out the UPS sounds like a reasonable idea. Can you thinkof anyreason that routing static IPs would make it more sensitive?
Unfortunately, I don't have a clue. :)Quote:
Originally Posted by hotfoot
Since you have experienced problems on numerous brands of switches, you might look at your cabling. Determine how long the cable drop is in ft. Also look at any interference that the cables may be subject to, track the cable routes and see if there is any access to intereference. If you are longer than 326' than all you need is a repeator on the line that's longer. If the cable is picking up interference, then you may have to shield the drop, or even worse, re-route the drop.
To determine the integrity of data is to put a sniffer in front of the switch and behind the switch.
When trouble shooting your network, you will save time and money by using the OSI model and trouble shooting it from layer 1 to layer 7. Always look at the Physical layer (layer1) first. Layer 1 consists of network interfaces, cables, hubs ect.
Since your problem only effects static IP's when you are running DHCP, I am assuming your printers, servers, copiers, ect., are the resources you are having problems with.
Switches normally operate on Layer 2, or the Data Link Layer, responsible for discarding bad data or incomplete data. This is where your cabling, interference and atenuation may come into play. The switch recieves the incomplete data and cannot forward it because the frames are incomplete. Do your switches lock up at the same time each day? If so Find out if there is anything happening that may cause intereference.
TCP/IP, IPX/SPX, and Netbuie all operate on levels 3 and 4, Network and transport layers(respectively). Level 3 handles addressing data packets for delivery, and also translates logical network adresses(IP) and Physical (MAC) adresses. Some switches that have built-in routing capabilities(WAN port) operate at the this layer.
I do not think your problem goes beyond layer 3 or layer two. I suggest you take a good hard look at your medium(cable) being used. But throwing different switches may be expensive (those 24 port Cisco switches probably set you back a couple K's each), and did not solve your problem. You need to go lower than switching to solve your problem.
Hope this helps.
Landisnet
http://www.landisnetusa.com