Re: %lineproto-5-up - down

Raiffeisenbank · ‎11-03-2004

Hi,

I have a really wired problem with our Cisco Network.

We have 3 4506 in the Core Layer and 28 3550 XL in the Access Layer. The Problem ist we have multiple Clients which have multiple offlines a day. The Time of the Offline is 5 Seconds due to 5 minutes. We get the following error on the Acces Switches, here is an example:

Nov 1 13:41:03 192.168.150.066 277:00:00 5w3d: %LINEPROTO-5-UPDOWN : Line protocol on Interface FastEthernet0/36, changed state to down

Nov 1 13:41:05 192.168.150.066 278:00:00 5w3d: %LINEPROTO-5-UPDOWN : Line protocol on Interface FastEthernet0/36, changed state to up

Nov 1 13:41:15 192.168.150.066 279:00:00 5w3d: %LINEPROTO-5-UPDOWN : Line protocol on Interface FastEthernet0/36, changed state to down

Nov 1 13:41:17 192.168.150.066 280:00:00 5w3d: %LINEPROTO-5-UPDOWN : Line protocol on Interface FastEthernet0/36, changed state to up

This happens on various machine on various times, there is not really a matrix in it ...

We did the normal testing like Cabling Speed an duplex

but nothing helps, should we autosense or fix the settings on all clients and Switchports ?

The Problem is really annoing, some of the Switches sometimes only react very slowly on telnet requests .. . dunno why.

We also have two multicast servers called vbrick which provide a mpeg stream for tv with 2mbits each.

Any ideas would be much appreciated.

Sample config of an access layer switch below:

! Last configuration change at 11:14:24 cest Wed Nov 3 2004

! NVRAM config last updated at 11:14:25 cest Wed Nov 3 2004

!

version 12.1

no service pad

service timestamps debug uptime

service timestamps log uptime

service password-encryption

!

hostname sw_w61_2og

!

clock timezone cest 2

ip subnet-zero

!

ip domain-name raibakwt.com

ip name-server 10.42.50.50

!

vmps reconfirm 1

vmps retry 5

vmps server 192.168.150.206 primary

vmps server 192.168.150.106

spanning-tree mode pvst

spanning-tree extend system-id

!

interface FastEthernet0/1

switchport access vlan dynamic

switchport mode access

no ip address

spanning-tree portfast

.

interface GigabitEthernet0/1

switchport trunk encapsulation dot1q

switchport mode trunk

no ip address

!

interface GigabitEthernet0/2

switchport mode dynamic desirable

no ip address

!

interface Vlan1

ip address 192.168.150.66 255.255.255.0

!

ip default-gateway 192.168.150.200

ip classless

ip http server

!

logging 192.168.150.206

snmp-server community <community> RW

!

line con 0

password 7 <password>

login

line vty 0 4

password 7 <password>

login

line vty 5 15

login

!

ntp clock-period 17180533

ntp source Vlan1

ntp server 192.168.150.101

end

Prashanth Krishnappa · ‎11-03-2004

Since you say you are seeing slow telnet response as well, check to see if the interfaces which are bouncing are seeing lot of broadcast or multicast traffic and errors. If you have multicast traffic, have you set up multicast routing and IGMP/CGMP in your network?

The following page should help as well

http://www.cisco.com/warp/public/473/46.html#maintain

Raiffeisenbank · ‎11-04-2004

We have Broadcast Traffic on the interfaces and some Multicast due to our streams and the normal network operation. We have no Multicast Routing at all. We Switch the traffic between our sites and dont route.

We did some checking with fixed speed and duplex settings, but im atm unshure if it solved the problem, ill have to wait until tomorrow to check the ifs again.

I heared there are some Problems with Intel Network Cards and the built in Asics in the switches...

I checked the telnet connections again today and it is definitly only one switch which has this problem, its a 3500 XL with 48 Ports, and the Multicasting Vbricks are Connected to it. But is a stream out of 4 mbit too much for the switch ?

Thanks for the article and the advices.

Raiffeisenbank · ‎11-04-2004

I checked our 4506 Core Switches as well today and found a really high amount of input errors and nearly all fibre gbit interfaces, i first thought of that bug:

http://www.cisco.com/cgi-bin/bugtool/onebug.pl?bugid=CSCec07464

but we are running IOS 12.1(20r) EW so we shouldnt be affected of this bug. And that means for me all our fibre links are broke... im really confused how to solve our major network problems that we have atm.

I replaced the switch which did answer slowly to telnet connections or dropped packets, but the new switch has the same problem without the multicast servers online ... I'll attach some logs and errors. I really somebody can help me.

Prashanth Krishnappa · ‎11-04-2004

You are running 12.1(19)EW1. 12.1(20r)EW is the ROMMON version and not IOS. So you are still vulnerable to CSCec07464. If you want to get past this bug, you might want to consider upgrading to 12.2(18)EW1

Raiffeisenbank · ‎11-04-2004

Hmm ok my fault... But as I read has the bug no affections on the performance.

As I wrote occurs the problem now on the replaced switch. It dropps some pings an answers radonmly slowly to telnet connections. If I make show proccesses I get max 18% utilization for 1 minute ... i dont know if this is much.

I solved the problems with the other clients by checking speed and duplex settings, but this one still makes problems with slow performance.