Connectivity issues with 3512 and 3508

[mcanet]

I had a lab setup that consists of 3 3512's that connect back to a 3508 via the GBIC ports. I had to relocate the lab, so I powered down the 3512's and installed them into the new facility. I confirmed that the fiber was terminated correctly so I connected the fiber to the 3512's and back in the closet to the 3508. Now I cannot get them to connect. There is no link light on either switch. I have tried different GBIC modules and differnet ports on both switches. Any idea how I can troubleshoot this or fix it? I reloaded the switch, I also copied over my backup config that I had, nothing works. I tried powering both switches down and bringing them back up, nothing. I have other switches connected to the 3508 and they work fine. Any ideas would be greatly appreciated.

 

[lui3]

Have you tried switching the fiber pairs on one side of the connection and seeing if the link comes up?


If that doesn work loop the transmit to the receive from one side of the connection and see if the light comes up if it does you know the configuration is correct and problem lies in the fiber between the switches?



Lui3
CCNP,CCDA,A+/Net+
Cisco Wireless Specialization

 

[mcanet]

Lui3, Tried that, tried different cables and switching around the transmit and receive. Luckily I have fiber cables with colored ends so that helps, but yes I did try this. Any other suggestions?

 

[lui3]

so when you looped the transmit back into the receive it did not bring the interface up?

ie take transmit patch cable and bring it back into the receiving tranmitter on the same interface

trans receive
u u
u u
u u
u u
u u u u

|
|
|
same cable strand

?


Lui3
CCNP,CCDA,A+/Net+
Cisco Wireless Specialization

 

[mcanet]

correct

 

[lui3]

i saw this once before but it was because the user was using a SM fiber patch when they should have been using a MM fiber patch. Can you verify that the fiber is MM on both the patch cable and the fiber run itself. you may need different gbics to get this to work along with different patch cables.

also when doing a show interface do you see the port state as being in err-disable?

Lui3
CCNP,CCDA,A+/Net+
Cisco Wireless Specialization

 

[mcanet]

Lui3,
No err-disable listed on the sh int. Also how would I even know if the cable is sm or mm? Is that something that would be written on the cable?

 

[lui3]

yes i on most cables it should say mm or sm. also whoever terminated your fiber run should also know? are you using the same patch cables that you used in the lab?

Lui3
CCNP,CCDA,A+/Net+
Cisco Wireless Specialization

 

[mcanet]

There is no metion of sm or mm on the cable. Same cables are being used that are used in the lab. Can I assume that sm = single mode and mm = multi mode?

 

[lui3]

i am wondering if the patch cables are correct but the fiber in the run is not the correct type. ie fiber run is single mode (sm) and the patch cables are mm (multimode).

hmm.

can you post sh run from 3508

Lui3
CCNP,CCDA,A+/Net+
Cisco Wireless Specialization

 

[mcanet]

fl-lab-3508-1#sh run
Building configuration...

Current configuration:
!
version 12.0
no service pad
service timestamps debug uptime
service timestamps log uptime
no service password-encryption
!
hostname fl-lab-3508-1
!
enable secret 5 $1$q272$AATGzxyvKra6WG8Pob4jL/
!
!
!
!
!
!
spanning-tree vlan 1 priority 16384
ip subnet-zero
no ip domain-lookup
!
!
!
interface GigabitEthernet0/1
switchport trunk encapsulation dot1q
switchport mode trunk
!
interface GigabitEthernet0/2
switchport trunk encapsulation dot1q
switchport mode trunk
!
interface GigabitEthernet0/3
switchport trunk encapsulation dot1q
switchport mode trunk
!
interface GigabitEthernet0/4
switchport trunk encapsulation dot1q
switchport mode trunk
!
interface GigabitEthernet0/5
switchport trunk encapsulation dot1q
switchport mode trunk
!
interface GigabitEthernet0/6
switchport trunk encapsulation dot1q
switchport mode trunk
!
interface GigabitEthernet0/7
switchport trunk encapsulation dot1q
switchport mode trunk
!
interface GigabitEthernet0/8
switchport trunk encapsulation dot1q
switchport mode trunk
!
interface VLAN1
ip address 10.10.2.1 255.255.0.0
no ip directed-broadcast
no ip route-cache
!
ip default-gateway 10.10.1.2
snmp-server engineID local 00000009020000024BC4EE00
snmp-server community fle-com RO
snmp-server community fle!!!wufsd RW
snmp-server chassis-id 0x13
!
line con 0
password tomcat
login
transport input none
stopbits 1
line vty 0 4
password tomcat
login
line vty 5 15
password tomcat
login
!
end

 

[lui3]

the only thing that has changed in this scenario is the fiber run being added. as long as you are using the same type of patch cables on either side and are using the correct mode of cable then it should work. the fiber run is either faulty or is using a different type of cable. that is my guess. would anyone know about the type of cable used to connect the switches from one end to the other. if they worked before in the lab and now do not work and configurations are the same then the problem has to be physical.

only other thing i can suggest is looking at the output from the

show logging

command and see if you can see what is happening



Lui3
CCNP,CCDA,A+/Net+
Cisco Wireless Specialization

 

[mcanet]

Ok, thanks for the quick replies. I have another question if you don't mind (if I should start a new thread just let me know). In the same building with the same hardware, I am having very slow speeds. The main switch in the building is the 2948, I am having trouble with the new and existing data drops. I have a Dell 2650 server that is connected to the 2948 at 100/Full. I tried to copy over a 100MB folder and data to a workstation that is connected to the 2948 via fiber and a media converter and it took about 20 minutes to copy over. I tried installing office xp from that same server and that is still going, I started it at 2:00 (now 3:15) Is there any troubleshooting utils or bandwidth monitors that I could put on the switch to see if it is having trouble?

 

[lui3]

On the 2948, what do you have the port speed set to. If the dell is set to 100 full the 2948 should be set to 100 full on the port connecting to the server. use the command

set port speed x/x 100

x being the module/port number of the port connecting to the server

this is typical of a misconfigured duplex issue


check the switch for errors using the

show port

command. look for FCS and Runts/Giants. Look at what ports they are occurring on. 100 or more is a lot. I don;t like to see any but sometimes i do let 3 or 4 slide. Normally they are associated with misconfigured duplex and port speeds on the hosts and switch ports. just make sure both sides match. auto/auto or full 100 and full 100.

you will know if the port on the switch is misconfigured if it is auto 100 half. this indicates the other side is set to 100 full.

this will make things go really slow.

check for the errors on the fiber side as well on both the 2948 and the pc if you can. i thought the 2948s used gbics for the fiber connections. mine do.

Lui3
CCNP,CCDA,A+/Net+
Cisco Wireless Specialization

 

[lui3]

oops, i didnt read that right. is your workstation connected via a 1000 base X fiber media converter to the gbic on the 2948?

sorry i misread the 100 mb to read 100 mbps.



Lui3
CCNP,CCDA,A+/Net+
Cisco Wireless Specialization

 

[mcanet]

The workstation is connected to a media converter at the room and in the closet. Then from the media converter in the closet cat53 into the 2948G. The Dell is set at 100/Full and so is the port it is plugged into. No erros, no runts, no giants either.

 

[lui3]

along with the port speed command you will also need to set the duplex to full on the port connected to the dell 2650


set port duplex x/x full  

this will result in the switch port connected to the dell 2650 being set to 100 full just like the server. this should fix the speed/duplex issue.





Lui3
CCNP,CCDA,A+/Net+
Cisco Wireless Specialization

 

[mcanet]

already was set to full on both switch and server.