We recently changed out our hub router at our data center and now I have no access to the management interfaces on the 2 netvanta 1550-48 switches there. These switches are stacked via gigax ports. They also have cross-connects into the data center colo. Multiple vlans are configured as well. The switches have been accessible since we installed them a couple years ago - up until now.
I'm fairly certain we've introduced a routing issue with the changeover of the hub router but the vendor cannot find it. We've spent hours with the data center networking team troubleshooting this as well with no luck. The cross connects go thru a Juniper vSRX firewall and connect multiple environments thru the stack.
An interesting twist to the story is: I had the data center tech do a power cycle on each switch while I had a continuous ping running. Pings started responding shortly after the first switch restarted. The pings stopped responding shortly after that. When the 2nd switch got power cycled the pings to it's management IP began responding long enough for me to log into the switch remotely. This lasted for maybe 10 minutes then I lost connectivity again to both.
The managment IPs for both switches are configured on a vlan interface. Does anyone have any idea where to look to fix this?
Oh and if it's not already complicated enough we're using Radius on the 2 data switches but that has worked flawlessly as well. Just cannot access them remotely at all.
Do you have any other IP interfaces configured on the switches?
If for example you have vlan 100 and vlan 200 both with IP addresses configured, and your computer is assigned to vlan 100, but are trying to access vlan 200's IP address, it will not work. You would need to access vlan 100's IP to gain access ot the switch.
This turned out to be a new piece of equipment that was sending hundreds of thousands of packets per second via the IPV6 protocol. It was compounded by the fact that a bug in the Juniper switching fabric at the data center was resending the packets even after unplugging the offending device.
We have since disabled ipv6 on the offending device and I can reach my switches again via their management IP's.