cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Announcements
139
Views
0
Helpful
0
Replies
Highlighted
Beginner

Possible Switch Arp-Cache Corruption?

Got a weird one - pretty much our largest customer has added a new building, and finally got fiber ran to it so I was setting up the WAN link between the new and old buildings.

For some reason, creating a WAN VLAN on each device, configuring ports as trunks, and then routing via static routes has not been working. Then today, the whole core just stops passing traffic, the only way to fix it was to clear arp-cache. Then, even weirder, after that happened no one can get out to the internet.

Core can see switch x that it is connected to, but cannot ping it. But switch y on the other side of the core can ping and ssh into switch x, with its only path to switch x being 
through the core. Resorted to reloading the core and am waiting to hear back if that fixes the issue.

The only configuration I made on the core was to create a vlan (yesterday) to act as the WAN connection, and a static route pointing to the subnet on the new building side. Again, all this was done last night and has been fine until this morning around 10 AM. They do have a router, which they dont know the password for, of course, which is why I have had to resort to plugging the new WAN link into the core layer 3 switch.

 

The core only has two static routes:

ip route  0.0.0.0 0.0.0.0 192.168.189.9 <-- to router/firewall

ip route 192.168.134.0 255.255.255.0 10.254.254.2 <-- new route to new building.  subnet does not exist anywhere else.

 

EDIT: The reload fixed the issue with connectivity, but the new building still has no connectivity to the core.

Everyone's tags (1)
CreatePlease to create content
Content for Community-Ad
July's Community Spotlight Awards