2960S Not Forwarding Traffic for Specific VLAN

[Terence Lockette]

Hey all,

I have a Cisco 2960S-PST-L 48 port switch that will either not forward traffic for one of our VLANs completely or it will forward a very small amount of traffic and then stop.  When I connect a laptop directly to the switch, it takes a while to pull DHCP information and my network connection icon (Windows 7) will either show as if it has a connection to the Internet or it'll show as if it has limited connectivity.  When I run ipconfig /all, I obtain all the necessary information needed to communicate locally and to the internet.  I can't ping by host name to a local machine or out to the Internet from my laptop.  From the switch I can ping out to the Internet via hostname, however.  If I put the switch port in a different vlan such as VLAN 1, I can get IP info and browse the LAN & Internet with no problems.  To ensure that it's not an issue with the DHCP server, Web filter, or Firewall, I've had a laptop tested at another switch of the same model and all works fine regardless of the VLAN; even the VLAN in question on this switch.  In troubleshooting, I've checked CPU history, STP, and interface errors.  No high CPU utilization, no loops, and no input/output errors, etc.  I have also deleted my vlan.dat file to wiipe out VTP, rebooted the switch, and let it re-learn VTP info.  It sounds software related and I haven't upgraded the IOS yet.  The current version of IOS is version 12.2(55)SE5.  The only change I made to this switch is adding all 48 ports to VLAN 60 on our network and removing the auto qos command from the interfaces as well as any lines added from using the auto qos command.  No other changes were made.  What could be the problem?  I can post a copy of the config if needed.

Regards,

Terence

[scheived]

You ever find anything more about this issue? I have a few 2960 stacks with the same issue, I reboot and they work for a few days.

[scheived]

I have an idea that it's caused by an invalid mac address on the network, I've disabled this port now and seeing if issue happens again.

-Nope still happening. The switch stops working pretty often now, maybe once every 2 weeks.

[scheived]

TAC was able to recreate this issue and assigned this bug, CSCvp9880 a DHCPV6 packets wedged in input queue buffer 3850.

[scheived]

I also found a device with a static ip address on the network same as the switch which would cause this. 

[musashi27]

@Terence Lockette did you ever find a solution to this issue? Was TAC able to solve the issue? Is it a Bug?

I have the exact same switch running the exact same IOS.

[Davis Oliveira]

Hi everyone!

I had a similar issue with:

Switch: WS-C2960S-48TS-L

IOS: 12.2(55)SE7

In my scenario, it's a golden jumper transporting a VLAN through a Port-channel (trunk mode - inside/server side) to a GigabitEthernet (access mode - outside/ISP side). Quite simple, but there wasn't communication between these hosts.

During troubleshoot, the server couldn't ping the ISP router interface and vice versa, although, this switch could learn both mac addresses of each side in the mac address-table.

I created an interface VLAN with a free IP address in the network, and from this switch I could ping both server and ISP router.

The ISP router didn't have any ACL, also.

Then, I noticed that this switch was the root bridge, so I changed manualy the root bridge priority to another switch (to a switch Nexus 7K behind the Port-channel, inside STP topology, which is the SPINE of the company).

This action resolved the issue.

But I didn't actually find any syslog pointing to a STP issue. Maybe it is a bug.

Still working on the RCA...