A few days ago we lost a 5548 switch in our data center. All interfaces that were supposed to be up showed up/up. The management IP address was pingable, could telnet to the box, no indications of resource problems with the switch itself. All SNMP metrics normal. First indication of a problem was that 2 hosts on the same subnet could not ping each other. Entries for those hosts were not in the ARP table. When entries were put in, hosts were able to ping. Saw no indications of local routing. Wireshark revealed no broadcasts present. Within a few hours, situation degenerated to the point that no servers on this switch were operational. All had connectivity issues. Logs, as far as I am able to discern, have no indications of serious or fundamental problems. This switch serves as a server access switch. Upstream is a 6509 acting as a server distribution switch. No indications of trouble there as well. Is there a possibility of problems on this trunk between these 2 platforms? Ready to replace the switch but concerned that this issue could happen again, as there were no error indications on our monitors for the 5548, or 6509. Any insight would be appreciated. Thanks.
