4500X - GLC-T ONLY works if cut down to 100 Mbps

[kylebrogers]

I have a new 4500X and a new GLC-T.  The 4500X recognizes when I insert and remove the GBIC.  However, the port its in will only change to UP if I set the speed to 100.  It refuses to come up if I leave it at the default.  I need to get the full 1000 Mbps out of the GLC-T.  Has anyone else run into this issue?  

You can see the status of the port below.  The moment I put "speed 100" on the port, 

Running 03.08.01.E code

show inventory:

NAME: "TenGigabitEthernet1/1/15", DESCR: "1000BaseT"
PID: Unspecified , VID: , SN: AGM2016419F

Note:  The GLC-T was ordered from Cisco.  The "PID: Unspecified" threw me for a minute but apparently different manufacturers (for Cisco) put in different info.

sh run int t1/1/15
Building configuration...

Current configuration : 99 bytes
!
interface TenGigabitEthernet1/1/15
switchport trunk native vlan 999
switchport mode trunk
end

<HOSTNAME>#sh int t1/1/15
TenGigabitEthernet1/1/15 is down, line protocol is down (notconnect)
Hardware is Ten Gigabit Ethernet Port, address is e00e.da7b.a78e (bia e00e.da7b.a78e)
MTU 1500 bytes, BW 1000000 Kbit/sec, DLY 10 usec,
reliability 255/255, txload 1/255, rxload 1/255
Encapsulation ARPA, loopback not set
Keepalive set (10 sec)
Full-duplex, Auto-speed, link type is auto, media type is 1000BaseT
input flow-control is off, output flow-control is off
ARP type: ARPA, ARP Timeout 04:00:00
Last input 00:27:10, output never, output hang never
Last clearing of "show interface" counters never
Input queue: 0/2000/0/0 (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 0 bits/sec, 0 packets/sec
527 packets input, 45530 bytes, 0 no buffer
Received 516 broadcasts (514 multicasts)
0 runts, 0 giants, 0 throttles
0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
0 input packets with dribble condition detected
1092 packets output, 87906 bytes, 0 underruns
0 output errors, 0 collisions, 3 interface resets
0 unknown protocol drops
0 babbles, 0 late collision, 0 deferred
0 lost carrier, 0 no carrier
0 output buffer failures, 0 output buffers swapped out

======================================

Here is what happens when I change the speed to 100:

<HOSTNAME>(config-if)#speed 100
<HOSTNAME>(config-if)#exit
<HOSTNAME>(config)#exit
<HOSTNAME>#sh int t1/1/15
TenGigabitEthernet1/1/15 is up, line protocol is up (connected)
Hardware is Ten Gigabit Ethernet Port, address is e00e.da7b.a78e (bia e00e.da7b.a78e)
MTU 1500 bytes, BW 100000 Kbit/sec, 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, link type is auto, media type is 1000BaseT
input flow-control is off, output flow-control is off
ARP type: ARPA, ARP Timeout 04:00:00
Last input 00:30:13, output never, output hang never
Last clearing of "show interface" counters never
Input queue: 0/2000/0/0 (size/max/drops/flushes); Total output drops: 0
Queueing strategy: fifo
Output queue: 0/40 (size/max)
5 minute input rate 2000 bits/sec, 2 packets/sec
5 minute output rate 5000 bits/sec, 5 packets/sec
677 packets input, 63048 bytes, 0 no buffer
Received 663 broadcasts (569 multicasts)
0 runts, 0 giants, 0 throttles
0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
0 input packets with dribble condition detected
1323 packets output, 108476 bytes, 0 underruns
0 output errors, 0 collisions, 3 interface resets
0 unknown protocol drops
0 babbles, 0 late collision, 0 deferred
0 lost carrier, 0 no carrier
0 output buffer failures, 0 output buffers swapped out

[Georg Pauwen]

There used to be a hidden command, 'service unsupported-transceiver', not sure if that is still available.

Which supervisor do you have ? I think there is a limit on the amount of Gigabit (4) and TenGigabit (2) ports supported...

[kylebrogers]

I gave the 'service unsupported-transceiver' a shot.  The command is still there, but it didn't fix the issue.  I think it recognizes it because it works at lower speeds.  

I'm using the 4500X as opposed to a 4500 blade chassis that has supervisor blades.  So it's just a 1U, 16-port, all 10G switch.  

I have 2 of them connected over a pair of 10G Twinax cables (running VSS so it acts like one switch).  Aside from those two twinax cables, the GLC-T GBIC is the only thing occupying a slot.  This is a new installation and isn't in production yet.  

[Georg Pauwen]

Hello,

I cannot find anything that points to any sort of limitation, but since your switches are not in production yet, I wonder if the port behaves the same in a non-VSS configuration, can you test that, to connect the port in a standalone setup ?

[Georg Pauwen · Accepted solution]

Hello,

looking at the problem again, and since your port stays down/down when set to 1Gbit, but up when set to 100Mbit, I am thinking that inadvertently you might be using a Cat5 and not a Cat5e cable, the first one being able to handle speeds up to FastEthernet, the second one being able to handle 1Gbit.

[kylebrogers]

I had tested it with 3 different cables just to make sure it wasn't a cabling issue.  After downgrading the code and having the same issue, I got a brand new CAT6 cable and it worked.  I don't have many CAT5(not e) cables.  It looks like I just had a bad string of luck in selecting my cables to test with.  

[Rytis Urnezius]

First of all i suggest to check cable, then changing GLC-T to another one, or test the same one on another device.

Finally i suggest to downgrade sw to cat4500e-universalk9.SPA.03.06.04.E.152-2.E4.bin (i'm running this one with GLC-T and i don't have any problems).

[Rytis Urnezius]

The same problem is discussed at this topic:

https://supportforums.cisco.com/discussion/12942111/1000base-t-glc-t-connected-sfp-interface-4500x-running-030702e-not-working

it confirms that GLC-T stopped working at 1gbps from sw >=03.07.xx