No MAC learned on port with Device attached- Up/Up

[Hans Schaefers]

Ok i am having an issue with some devices plugged into a 2960 switch where the switch is not learning the MAC address of the devices attached. Has anyone ever had this issue and what was your solution?

• Plugged a TPlink 5 port into the 2960 and the device into it and the 2960 learns the MAC attached through that small switch but when connected directly - nothing. So it will learn the mac of the device and the device does work on any other non Cisco switch as I did test it with several.

MainSwitch(config-if)#do show mac add int gi 1/0/2
Mac Address Table
-------------------------------------------

Vlan Mac Address Type Ports
---- ----------- -------- -----

!
interface GigabitEthernet1/0/2
switchport mode access
spanning-tree portfast
!

MainSwitch#show int status
Port Name Status Vlan Duplex Speed Type
Gi1/0/2 connected 1 a-full a-100 10/100/1000BaseTX

MainSwitch#show int gi 1/0/2
GigabitEthernet1/0/2 is up, line protocol is up (connected)
Hardware is Gigabit Ethernet, address is 001a.e365.c682 (bia 001a.e365.c682)
MTU 1500 bytes, BW 100000 Kbit, DLY 100 usec,
reliability 255/255, txload 1/255, rxload 1/255
Encapsulation ARPA, loopback not set
Keepalive set (10 sec)
Full-duplex, 100Mb/s, media type is 10/100/1000BaseTX
input flow-control is off, output flow-control is unsupported
ARP type: ARPA, ARP Timeout 04:00:00
Last input never, output 00:00:01, output hang never
Last clearing of "show interface" counters 2d09h
Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0
Queueing strategy: fifo
Output queue: 0/40 (size/max)
5 minute input rate 0 bits/sec, 0 packets/sec
5 minute output rate 3000 bits/sec, 3 packets/sec
11589 packets input, 7103264 bytes, 0 no buffer
Received 9172 broadcasts (21 multicasts)
0 runts, 0 giants, 0 throttles
0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
0 watchdog, 21 multicast, 0 pause input
0 input packets with dribble condition detected
329940 packets output, 47254573 bytes, 0 underruns
0 output errors, 0 collisions, 0 interface resets
0 babbles, 0 late collision, 0 deferred
0 lost carrier, 0 no carrier, 0 PAUSE output
0 output buffer failures, 0 output buffers swapped out

[Mark Malone]

have seen macs not been added before for various reasons , looks like the interface is seeing broadcast packets , what are the devices you are connecting that are having the issue , does it happen with standard devices pcs/phones etc

is it happening with all connected devices or only certain types

if you hardcode the mac to the switch does it see it ok ? static mac

what software version is running on the switch 

do you see anything show logs , is spanning tree showing fwd for the ports connected

[Hans Schaefers]

Toanswer your questions,

• I am trying phones, ata's, pc's (desk and laptops), as well as other switches.
• Appears to only be pc's, phones and ata's.
• Tested and setting a static MAC address in the table does not have an affect.
• c2960s-universalk9-mz.150-2a.SE9.bin

Some additional notes:

• Other switches work fine and I understand that is because they detect each other through the BPDU's seen.
• I have tried pc's with win10,7 and ubuntu 14lts all have same results.
• I have tested phones and ata's from cisco, grandstream and yealink with the same results.
• The switch has been factory reset and everything is on vlan 1 then I repeated all of the tests with the switch as the DHCP server and completely isolated.

So knowing that other switches will negotiate a trunk port, i configured 1 interface int gi 1/0/1 as a trunk port. Enabled portfast and nothing else.

Now every pc, phone or ata I plug in gets a DHCP from the switch it self.

I disabled the switch DHCP server and plugged it into my test network. The devices still pull an IP.

Let me ask this question now:

Why does a port need to be set to Trunk for an access type device that only needs to access the default vlan or the NON tagging vlan on the interface?

It almost feels like packets with tags are being sent and or the latest code is not accepting any packet. I have see interfaces receive 1 type of packet before for a specific vlan and would get Vlan Mismatch errors. Mind you i have usually seen this when servers were set to trunk on the VM and I was instructed that they only need one certain VLAN and need to manually prune that.

I did run a debug to identify vlan tagging on the access port and get nothing.

I'm open to any ideas.

[Mark Malone]

If you plugging a phone and pc directly in the 2960 and your not learning any mac addresses but they work fine on other switches there may be hardware issue , your on the best software release for that switch currently

Why does a port need to be set to Trunk for an access type device that only needs to access the default vlan or the NON tagging vlan on the interface

They should not be set as trunks if there access devices unless there acceptimng multiple vlans

on pc/phone interface

you set

switchport access vlan 10

switchport voice vlan 11

On trunk only to be used between switches or connections to devices which use multiple vlans like  a san

switchport mode trunk

Never use portfast on a trunk connection either side

......

either way configuration should have no effect on a switch learning a mac from a directly connected device , once the nic is online the mac broadcasts out to the switch port and should be learnt in , layer 1 and 2

Do the ATAs show up in show cdp ? do they not show up in the mac table either locally ?

Are you saying setting a static mac doesn't show up in your mac table ?

Have you tried to span any of these ports with a wireshark to see whats happening at the wire level when a device connects  ,if the switch is seeing the broadcast or refusing it

Has this switch worked before fully there always the chance of hardware issue with it if it wont learn anything directly connected into it

Now every pc, phone or ata I plug in gets a DHCP from the switch it self.

Where is the arp taking place whats the layer 3 device, are you seeing ,macs resolved to the ips there ?

[Hans Schaefers]

"Never use portfast on a trunk connection either side"

I agree, but without portfast the devices were not seen. Its odd I admit.

Do the ATAs show up in show cdp ?

Yes when in trunk state. Both Cisco and Grandstream ATA are factory reset so they are running default settings.

Are you saying setting a static mac doesn't show up in your mac table ?

no, a static MAC does show up in the table when set, but setting a static MAC had no affect on the Access Port function when I had it programmed as an Access port  - access vlan 1

Have you tried to span any of these ports with a wireshark to see whats happening at the wire level when a device connects  ,if the switch is seeing the broadcast or refusing it

Yes I have tried to span the port and I am seeing what appears to be regular boardcast for DHCP requests. The switch is not responding. I have checked my trunk also and it is not relaying DHCP from any port either for directly connected devices.

Has this switch worked before fully there always the chance of hardware issue with it if it wont learn anything directly connected into it.

Yes, it has worked before. I have 3 more switches at the office that are showing similar issues. The 2960 is an older switch so we are planning on replacing them and I am not having this issue on the newer ones. I just don't understand the malfunction and its been bugging me.

Where is the arp taking place whats the layer 3 device, are you seeing ,macs resolved to the ips there ?

In this test environment, this is a flat topology. connected devices are the L2-L3 devices. There is no routing between each device such as DHCP server, Phones, Phone Server (Old BE6K) or AP's. Just 3 switches. HP Procurve, Cisco 2960, Meraki MS220-8P.
No when the 2960 is set for access port for vlan 1, MAC is not seen on any other device directly connected to the switch. Its like Microsegmentaiton or something is dropping it before repeating to the other interfaces. This sounds like a hardware issue, but to have a total of 4 devices that started doing it at the same time with no change in the environment is really odd.

[Mark Malone]

Bit odd alright that 3 would go at the same time , anything in common with these 3 switches are they running the same IOS even though your on the safe harbour version have you tried changing the software on 1 of them in case a bug got triggered somehow just to rule out the software

The fact a device connects direct to the switch you cant see the mac would usually suggest a nic issue on device as the switchport learns it in by broadcast but if its happening with multiple devices on same switches and these devices work fine on other 2960s it has to be software or hardware doing something irregular

I understand your saying its a flat layer 3 network but there has to be a layer 3 device there somewhere that resolve macs to arp , the core switch or break out router , does that resolve any of the macs to the ips that don't show up on the switch or does it show them as incomplete , Im just wondering if you set as a test one of the devices to be static mac and static arp would it show up as working it could indicate an issue with the switch learning things dynamically but still working when set manually