04-22-2003 04:38 PM - edited 03-02-2019 06:48 AM
Not really sure it is correct to call it a flapping port, but the Gig port at random intervals changes state to up. The message is:
3d20h: %LINK-3-UPDOWN: Interface GigabitEthernet0/2, changed state to up
It never reports the interface or protocol is going down. The light changes from green to amber, pings die for about 30 seconds consistantly. This disturbing behavior occurs every 3, 5, to 30 minutes. No pattern as to when it happens.
This is a new switch I am trying to put in service.
I have changed the GBIC, cleaned and inspected the fiber, and swapped ports without changing the problem.
I have turned off trunking on both sides (switchport mode access). Autonegotiation selects 1000 Mb and full duplex on both ends of the fiber.
It sort of looks like the port is randomly rerunning autonegotiation. Show Int number look good - I pasted it below.
After everyone goes home, I'll swap out the switch. This seems like pilot error, not something broke. Any suggestions?
show int g0/2
GigabitEthernet0/2 is up, line protocol is up
Hardware is Gigabit Ethernet, address is 000b.fd80.671a (bia 000b.fd80.671a)
MTU 1500 bytes, BW 1000000 Kbit, DLY 10 usec,
reliability 255/255, txload 1/255, rxload 1/255
Encapsulation ARPA, loopback not set
Keepalive set (10 sec)
Full-duplex, 1000Mb/s, link type is autonegotiation, media type is SX
output flow-control is off, input flow-control is off
ARP type: ARPA, ARP Timeout 04:00:00
Last input 00:00:00, output 00:00:00, output hang never
Last clearing of "show interface" counters 23:49:04
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 105000 bits/sec, 111 packets/sec
5 minute output rate 105000 bits/sec, 111 packets/sec
574197 packets input, 70795070 bytes, 0 no buffer
Received 283014 broadcasts, 0 runts, 0 giants, 0 throttles
0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
0 watchdog, 40977 multicast, 0 pause input
0 input packets with dribble condition detected
108456 packets output, 12818992 bytes, 0 underruns
0 output errors, 0 collisions, 2 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
04-22-2003 10:55 PM
The line up messages without down matched messages is a problem first found in 8500 platform. You need check the remote peer end of the giga interface. Here is DE's explanation for your reference:
85xx (or any other IOS platform) can report multiple link-up messages without
a corresponding link-down message when it gets the interrupt for link-up from
the peer. In other words, if the peer goes down for a very short time and comes
up it (peer) sends a link-up interrupt. 8540 receives this and finds that the
link is already up and in response to the link-up interrupt it prints a message
on the console indicating a link-up event. This is not an issue, and in most
cases are genuine - this is especially during the bootup stage, when 8540 does
auto-negotiation. Sometimes 8540 comes up (after auto-neg) and the peer had not
come up, and the peer starts sending link-up interrupts multiple times. These
are the cases where we see 8540 reporting multiple link-up message.
09-15-2003 03:05 PM
I am dealing with the same problem here! Does the interface on your switch goes into errdisable state? The interface on my switch goes into errdisable state and I have to reload "bounce" the switch or manually do a shut and no shut on the port to bring it back! It could be a bug on the Cat 3550 series switch. The next time it happen I am sending cisco the show tech and leting the expert deal with it.
Mark,
09-15-2003 11:32 PM
Hi Mark,
There are a couple of commands that can help you determine why it went to errdisable state - show errdisable detect, show errdisable recovery. Show log may also have something of interest. These can definitely help when sending the show tech to TAC.
-Robert
09-17-2003 07:22 AM
Robert,
First of all thanks for you reply. I had been doing errdisable detect and recovery int the last few days already. The link only goes down when there are lots of traffics on the line and I had try swap GBICs, ports, fibers and I even put a new switch in and all of them didn't help. The link only errdiable when the link flap some many time in 10 seconds and I had errdisable detect and recovery enable and waiting to see what the log is going to tell me this time.
Mark,
09-17-2003 09:14 AM
This looks like CSCea87747, which is fixed in 12.1(14)EA1.
Problem Description
A Catalyst 3550-12G logs the following syslog message on a semi-regular basis when
connected to a Catalyst 8500 Gigabit Ethernet interface
%LINK-3-UPDOWN Interface GigabitEthernet0/9, changed state to up
This problem has also been observed when connecting a Catalyst 3550 to a WS-X5550
or WS-U5534.
Conditions
The Catalyst 3550 is running 12.1(11)EA1 or above.
Workaround
Downgrade to 12.1(9)EA1.
Additional Information Regarding Fix
When running 12.1(14)EA1 or later, both the 3550 and link partner (8500 in this case)
must be configured for auto-negotiation.
________________
Now although the bug describes an 8500 as the link partner, this same behavior has been seen on the
5500 and 6500 switches as well.
09-17-2003 09:20 AM
I thought there was a problem with spanning tree bug with ios version 12.1(9)EA1, where the switch sending out bpdu to the wrong MAC address.
Mark,
09-15-2003 11:25 PM
Hi,
The reason the interface does not report down is documented in bug CSCdz00065. It is fixed in version 12.1(13)EA1. I recommend to run the latest release for stability, 12.1(14)EA1. You may want to try swapping out the fiber, than GBIC, or move the link to the other Gig port to see if the problem follows.
-Robert
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide