05-08-2008 06:19 AM - edited 03-05-2019 10:50 PM
6513 running CATOS Version NmpSW: 8.3(2) and IOS Version 12.2(14)ZA7. Switch had been up for 2+ years. Switch hosts GRE tunnels to other offices. Packet loss down all tunnels was noticed. Company had seen this in the past and a reboot would fix.
Rebooted 6513 no errors in the reboot. But my Vlan20 that is layer3 had issues. The serves on Vlan20 could not ping their default gateway. But all other servers could ping their L3 Vlan on the switch.
The "fix" was to "no interface Vlan20" on the IOS part and past it back in. Now the servers on vlan 20 10.122.20.0/24 could be seen by all L3 vlans and they could ping their dg of 20.1.
Also, the all L3 vlan interfaces are encountering packet drops. If any pc pings its L3 Vlan packet drops occur. Example if pc on vlan 24 pings its 10.122.24.1 dg, packet loss. Also if same PC ping 10.122.20.1 L3 Vlan on 6513, packet loss.
However if a pc on Vlan24 pings another pc on another Vlan (Vlan20) No packet loss occurs.
05-14-2008 06:02 AM
You can go through this command after hours:
shut
no shut
on the VLAN interfaces you are having issues. this should clear the cef tables entry.
05-14-2008 07:13 AM
Thanks. I have never seen a hybrid do that. How much more "clear" can a CEF table get after a power off/on of a switch? :)
Spoke to a cisco tech who verified that cisco has seen this behaviour before, but doesn't have any type of fix for it. I hate hybrid mode, would love to move it to native but we all know how painful that is.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide