Teneo !!!

Aaron’s blog on Networking, and Enterprise Technology

Metro Ethernet Frustrations and the fix

For those that have been following the last few weeks, you’ll know that I was implementing a Metro Ethernet (MAN) network for my company using a Cisco ME-3750 switch, and the frustrations that came with it.

Well, I still went forward using the ME-3750, just not using the ES Ports.  The circuit came up fine.  I thought I was done.  Ran a few simple tests…. yep, communication is working.  So I left to the airport.

A couple of days later, communications started failing.  Intermittent results.  Basically, it stems from ARP not working correctly.  I would try to ping a device, then look at it’s ARP table.  Nothing….. hmmm… wierd.

I then look at the Switch’s mac-address table… yep…. the MAC’s are populating.  So, I go the distance, and place two network sniffers on each end.

The end result, was that sometimes (and only sometimes), an ARP request would get sent out, but the reply would never com back across the link.  Since I was monitoring the actual trunk ports, this must be a problem with the provider.

Well, come to find out, AT&T (who was doing the fiber to copper hand-off for me) has a mac-address table with a maximum entry of 50 entries.  50 entries!!  We have a decent sized network, and we are moving servers to a co-location DR site.  C’mon… we’ll max 50 entries in no time.

I didn’t even know AT&T caps mac-addresses.  Do other providers do this, as well??

Technorati Tags: , ,

2 responses to “Metro Ethernet Frustrations and the fix

  1. Sean February 10, 2008 at 11:50 am

    Any time I’ve got a LAN extension service we’ve been required to route across it, no bridging. I didn’t realize you had one, for some reason I assumed it was a direct connection between the sites.


  2. Aaron Paxson February 10, 2008 at 1:40 pm

    Hi Sean. Yes, that was my original design. However, due to the original design of the network (there wasn’t one), and due to the IP Address schema, and the fact that most clients are using static IP’s and not DNS, I had to bridge across the link.

    Personally, I cringe at all the broadcasts going over this fiber link.

    We are moving servers to the remote site for DR. But, since most clients are using static IP’s for some of the servers moving over, I didn’t have much of a choice. That is something that will be resolved eventually, but that’s a long-term goal.

    **EDIT** Now that I think about it, that explains the MAC-limitation. Usually, when you route over this interface, you are only looking at the MAC’s for the routable interfaces, but since I’m bridging, there are so many more. Yeah, that makes more sense.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: