Jump to content

Recommended Posts

Posted

Hi,

 

We have a point to point connection between two sites which has recently gone down and the ISP have told me the issue is that there is a loop in our internal network causing the issue as their layer 2 devices are learning Mac addresses. My internal config has not changed so I am not sure how a loop has suddenly occurred.

 

I would like to verify that the point to point connection actually works by plugging in a device in one end and another end but not sure how to test if they actually will communicate.

 

Any help would be much appreciated.

Posted

Plug a laptop in either end and see if they can ping each other

 

It is a bad idea to have a Layer 2 connection on a WAN link. The data should be routed.

 

The fact they are learning Mac addresses doesn't indicate there is a loop. I would expect that if you're using the connection as a layer 2 link.

Posted
Usually I would ping an IP address but as a layer 2 device does not give out an IP address, should I ping the Mac address?
Posted

You can't ping a Mac Address.

 

Layer 2 switches can I have pingable IP address so you can ping that.

 

Or Ping an IP of a server or something on the other side.

 

To have a proper test you should do the laptop thing above. Test it with a basic switch first and then plug them directly into the ISP equipment and try to ping.

 

What type of switches do you have on each side?

Posted

Apologies, i should have been clearer.

 

The switch that the ethernet cable runs from to the ISP`s switch is a 8206ZL and on the other site a ethernet cable runs from the ISP`s ethernet switch to a HP Switch 2530

Posted
If they're correct then a loop would most likely be someone's plugged both ends of an ethernet cable into 2 network points on the wall in a random room on one of your sites
Posted

From what you are saying this is a LES (Lan Extension Services) circuit which does act as a layer 2 point to point link.

 

 

If you did have a switching loop I'd expect to see more issues on the LAN not just the LES link. Could you try different cables between LES switch and the HPs.

 

If you do discover a loop you could isolate bits of your network one by one to find the specific switch, the port activity lights on the switches normally go mad if there is a loop from the one time I have witnessed this issue. If they are managed switches with console ports you may also find a biot more info from the cli.

Posted

Hi,

 

 

The switches are managed so what command should I run to identify the loop. THere is a switch that is blinking faster than the other so for the time being as a test, if I was to power this switch down, would this resolve the issue?

Posted (edited)

Jus running doing a

show logg -a
could show duplicate MAC addresses/conflicts.

 

The only one time I have seen this is where different vlans were used and the only devices effeected were devices on that vlan i.e. Internet Browsing, shared drives etc. It was quite easy to spot something was wrong by the port activity lihgts on the switch, it jus did not look normal. It turned out to be someone had put a network cable between two wall ports.

Edited by Davit2005
Posted
Also where I've had a loop before it was so demanding on the switch I couldn't even connect to its management interface, easily told me which switch was causing the problem. Admittedly it was about 10 years ago on an old 3com 4400 stack.
Posted

Finding loops is a pain but start by working through the problem methodicaly

 

Reboot the switches and see if that fixes it.

 

Do all lights flash together

 

Remove all connections from switch on none working end and check working end switch is working.

 

Now only connect cable that connects to working switch reboot switch and check it's working with laptop.

 

Now slowly plug other cables in till it stops working giving some time between pluging next cable or switch in.

 

You might find a loop or a bad nice card that's broadcasting flooding switch

Posted

After doing a lot of testing it turns out we do have a loop but not on the site where the internet has gone down but on the other site where the internet is up. I have had reconfigure the link between the two sites on another port. It seems that whenever I plug into the original port the port on the other site gets disabled because of STP, so the loop is on the other site.

 

Seems I will have to unplug cable by cable as it is a mess here.

Posted

You have already split the problem in half. Just keep going, and you will isolate the loop in a very short period of time. Read the linked article!

 

Though if spanning tree is taking out ports, it may be a spanning tree/port config issue.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...