what could cause interface up/downs?

[NettableWalker]

I have installed a new Cisco 3560-48 and the logs keep showing interfaces flapping from time to time.

What could be causing this? is it likely to be the PC or the switch?

Is there any further troubleshooting i can do to get to the bottom of it?

Jun 8 05:33:39.976: %LINEPROTO-5-UPDOWN: Line protocol on Interface FastEthernet0/12, changed state to down
Jun 8 05:33:40.974: %LINK-3-UPDOWN: Interface FastEthernet0/12, changed state to down
Jun 8 05:33:43.608: %LINK-3-UPDOWN: Interface FastEthernet0/12, changed state to up
Jun 8 05:33:44.615: %LINEPROTO-5-UPDOWN: Line protocol on Interface FastEthernet0/12, changed state to up
Jun 8 05:33:47.887: %LINEPROTO-5-UPDOWN: Line protocol on Interface FastEthernet0/12, changed state to down
Jun 8 05:33:48.885: %LINK-3-UPDOWN: Interface FastEthernet0/12, changed state to down
Jun 8 05:33:51.217: %LINK-3-UPDOWN: Interface FastEthernet0/12, changed state to up
Jun 8 05:33:52.223: %LINEPROTO-5-UPDOWN: Line protocol on Interface FastEthernet0/12, changed state to up


MCP,CCA, Net+, 1 quarter CCNP...

 

[vipergg]

It could just be normal users bringing pc's on and off the network . Also could be caused by speed/duplex mismatches between the switch and the users nic cards , you would have to check the ports for errors to determine that..

 

[NettableWalker]

Hi Vipergg, thanks for the help.

I've since moved the user to another port and the same is happening so i reckon it's a fault with their PC. If in doubt I blame windows over cisco everytime!

MCP,CCA, Net+, 1 quarter CCNP...

 

[baddos]

It could be some radical interference on the cabling or a cable that is beyond the Ethernet distance limitation.

 

[enzo288]

That is a good philosophy to blame Windows over Cisco.

CCNA, CCDA, CCNP, CCDP, almost CCIE

 

[jneiberger]

It could also be a bad cable, a bad data drop, a faulty NIC...

 

[NettableWalker]

I think it was probably duplex/speed settings being set ot auto. As soon as i changed everything to 100/full the problems went away. Strange as these were HP desktops and only 1-3 years old mostly.

MCP,CCA, Net+, Half CCNP...

 

[jneiberger]

Duplex and speed settings on a link will not cause it to bounce. If it's bouncing, the problem is electrical or physical. Something is wrong with the NIC, the cable, the data drop, the jack on the NIC, the cable connectors...whatever. Misconfigured speed and duplex settings will cause suboptimal performance but they will not cause the link to bounce by themselves.

 

[NettableWalker]

They weren't really bouncing, just going up/down randomly perhaps every 2 minutes or 30 minutes.

The problem has definitely disappeared now.

Now I have a much bigger one:

Picture a gigabit fibre SFP on a 3560 going straight to a gigabit SX module on a 2950.

Native VLAN dot1q trunk.

pings and network connectivity are fine but browsing the network was v.slow and a 50 MB file transfer would take 25 minutes.

So much for my "upgrade" any ideas what could possibly cause this? I have temporarily pulled the 2950 and re-installed the very old hub. and things are functional again.



MCP,CCA, Net+, Half CCNP...

 

[jneiberger]

Were you seeing any errors on either interface? I wonder if you had a bad fiber connection or cable. Do you have a way to test it?

 

[NettableWalker]

I can only test for actual connectivity across the fibre line (a red light) and no errors were visible at all.

would an Ethereal trace help me in any way. Is there a type of traffic that would appear in this case?


Very strange, don't you think.

MCP,CCA, Net+, Half CCNP...