tobyhansen
asked on
Problems with Cisco switch and some of the ports
We have a Cisco 3650 switch with a flat vlan that is linked to another Extreme switch via a fiber whip and mini-gbic. When I got to the office it appeared that the gbic on the Cisco switch had died and when I replaced the gbic the link came back up. Well when it appeared to be fine again I noticed we started losing connectivy to a couple servers, one being a VM on a Hyper-V server. You could get to the Hyper-V but not the guest. odd. Anyway it started to progress and then it was a handful of servers that could not communicate. I immidiately unplugged the uplink and the problem was still there. We traced the issue down to 2 ports that seemed to be bad. For example - When we hung a workstation off the port to test we could hit a some addresses of servers on the switch and others we couldn't. Very inconsistant. Since the issue yesterday I have left the other switch unplugged and we are not using the 2 (so called) bad ports.
What I am trying to find out is if there is a correlation between the port that failed and the issues I had with the intermittent connectivity with just SOME of the ports and VM's on the switch. I can explain this a little deeper if needed but was wondering if this may hit a nerve with someone who has had a similar experience.
Again seemed to be when I plugged the switch back in that the problems started to creep in. It wasn't all at once but progressive. Even after I unplugged the uplink the problem was still there.
What I am trying to find out is if there is a correlation between the port that failed and the issues I had with the intermittent connectivity with just SOME of the ports and VM's on the switch. I can explain this a little deeper if needed but was wondering if this may hit a nerve with someone who has had a similar experience.
Again seemed to be when I plugged the switch back in that the problems started to creep in. It wasn't all at once but progressive. Even after I unplugged the uplink the problem was still there.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Do you have multiple NICs connected to the same switch from the Hyper-V host server? Depending on how the virtual switching is setup, this could create a spanning tree loop.
Also, I have seen broadcom network cards create loops and steal IPs and all other sorts of non-sense when they are teamed using the Broadcom software and then used for VMware/Hyper-V/Xenserver main network links.
Also, I have seen broadcom network cards create loops and steal IPs and all other sorts of non-sense when they are teamed using the Broadcom software and then used for VMware/Hyper-V/Xenserver main network links.
ASKER
It could have been something with the Hyper-V server as that was the machine that had the original issues. It was very odd. You could get to the host but not 'some' of the VM's. The thing is I think they are all on the same virtual switch with just one NIC going to the physical switch.
Can these issues be progressive? Can they start small and escelate for no reason to other servers, vitual and physical?
Is there a way to tell on a Cisco switch exactly what ports may have been flagged by STP and limited or blocked? I could not find any evidence in the switch logs. I am about to turn this back on in the morning.
Can these issues be progressive? Can they start small and escelate for no reason to other servers, vitual and physical?
Is there a way to tell on a Cisco switch exactly what ports may have been flagged by STP and limited or blocked? I could not find any evidence in the switch logs. I am about to turn this back on in the morning.
Yes, you can turn on debugging to watch the STP events.
switch# term mon
switch# debug spantree events
Might be slighlty different debug command I have an older switch with older IOS on it.
switch# term mon
switch# debug spantree events
Might be slighlty different debug command I have an older switch with older IOS on it.
SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
We ran this and didn't see any blocked ports. Yet now we got another port that seemed to go haywire.
We can ping just a couple other servers on the switch when connected to one of these failed ports but can't hit anytihg else. Very odd. Any thoughts?
Again the server on the port that goes down can no longer communicate out... except for just a couple other ports. STP doens't show any blocked ports. Now we have 3 ports that when we plug into can only see a couple IP's.. and nothing else. Not sure if I am explaining the right.
We can ping just a couple other servers on the switch when connected to one of these failed ports but can't hit anytihg else. Very odd. Any thoughts?
Again the server on the port that goes down can no longer communicate out... except for just a couple other ports. STP doens't show any blocked ports. Now we have 3 ports that when we plug into can only see a couple IP's.. and nothing else. Not sure if I am explaining the right.
any security on the switch? Can you post a sanitized version of the config?
ASKER
Here you go... like I said very basic.
switch.txt
switch.txt
That's about as basic as it gets.
could just be a bad switch.
You could check the arp table on the switch or the mac-asddress table see if something looks weird but that config is almost default.
could just be a bad switch.
You could check the arp table on the switch or the mac-asddress table see if something looks weird but that config is almost default.
ASKER
We reset the switch and the problem so far has been eliminated. Thanks for the help.
ASKER