Intermittent Vlan subnet connection

nojpt · ‎11-23-2006

Hi,

I have a very strange problem on our network. we have about 72 vlans configured on our 6515 core switches that acts as distribution and core. Each vlan has different /23 subnet and which are distributed to 3750 access switches. These access switches are uplinked to the core switches via fiber trunk lines.

For no reason at all, or at least no logged messages anywhere, one vlan subnet just can not connect anywhere. The PC's connected to this access switch can obtain correct dhcp address but cannot ping anything including the gateway.

Any help will greatly appreciated. Config pertaining to that vlan is attached.

Any help would be greatly appreciated.!

sourabhagarwal · ‎11-23-2006

if PC is getting correct IP and it is able to ping gateway, it means request is reaching till core switches.

can you check routing on core switches for 172.23.88.0/23 subnet?

if PCs in 172.23.88.0/23 subnet are not able to ping PCs in other VLAN which is created on core switch, then it is an issue of inter-vlan routing on your core switches.

i would suggest you to first check routes on core switches.

let me know if it helps or you have any question.

nojpt · ‎11-23-2006

Hi,

Route 172.23.0.0/16 is connected subnets are all my interface vlans are within this subnet. So this summary route is considered connected routes. See below result of 'show route'

C1#sh ip route 172.23.0.0

Routing entry for 172.23.0.0/16, 84 known subnets

Attached (80 connections)

Variably subnetted with 4 masks

Redistributing via ospf 1

I don't think this is a routing issue but rather hardware or software bug.

We actually created another vlan 164 and assign a new subnet (172.23.164.0/23) to this new vlan and have the users get dhcp address from this subnet. It only run for 1 week, then the problem of loss of ip connectivity recurs. Same problem, they can lease ip from dhcp server but unable to reach anywhere outside this bridge.

amit-singh · ‎11-24-2006

One of my customer faced the same issue on 4500 switches. We had a server vlan interface that use to went off once in a while.

The way we use to make it up was using a "shut and no shut" on the vlan interface or rebooting the switch. This was a cosmetic issue with the IOS that we were running. There was not any bug reported as such. We upgraded the switch to the latest IOS in the train and everyting is working fine since then.

I am not sure if that could be your case but I just wanted to share my experience on the post.

HTH,

-amit singh

nojpt · ‎11-24-2006

Thanks for sharing your experience. We used to do that too. Shutdown the interface vlan on the active router so the other router on HSRP takes over. Traffic will be ok for a few minutes then off again.

The new interface vlan workaround actually worked for 1 week until the same problem recur today on the same access switch.

I'll gather some more responses here before i decide for IOS upgrade.

glen.grant · ‎11-24-2006

When this happens can the pc's ping each other on the same vlan ? You say they pull addresses , if they are on the same vlan they should be able to ping each other which would then move you up to the core switches to look around otherwise it is local to the 3750's .

nojpt · ‎11-26-2006

They are able to ping pc within the vlan and they are able to pull ip address from dhcp server which is in another vlan. But they are not able to reach anything else. Strange.