I'm curious to hear from anyone out there who has more experience than I do working with T3's for data WAN connectivity (which should be most of you). I'm testing out an MPLS WAN link for a customer.

It's a pair of T3's on either end throttled for 15mbps of throughput. I'm using Wankiller to stress it. I'm finding I get some errors, though not a lot (see below). Not having a lot of experience with T3's I don't have a good sense of whether this is within spec or something I should bark at ATT about before going live with this thing.

Serial #/# Serial#/# is up, line protocol is up Hardware is DSXPNM Serial Internet. Show interfaces Serial. There are no errors on the T3 controllers or the serial interfaces and yet I am unable to generate more than ~21Mb/s traffic using netcps. Both routers are running IOS 12.3(9). Here is the T3 controller and serial interface config from our routers. The only difference between the. Hardware is DSXPNM Serial. Used for communication between the Arduino board and a computer or other devices. All Arduino boards have at least one serial port (also known as a UART or USART. How can I check the traffic on a serial interface to be sure that it is not overloaded? I having an issue with an application not working over our WAN connecti. Hardware is DSXPNM Serial Description: DS3 to Wilmington. Do you have the interface stats and config from the other side as well?

The counters below show what I've seen over about 25 minutes of testing, pushing a little over 13mbps of traffic across it. It used to be fairly common for parts of a T1 or T3 to be mis-configured within the telco cloud, as voice circuits used to use a different line encoding than data circuits. This would cause errors on certain strings of data; ie, if a packet contained eight 0s in a row, etc. You can try to ping from one end to the other with certain patterns and try to see if one will have issues (such as all zeros, all ones).

Its easy to do this from an enable prompt ('ping 12.112.7.173 data ffff size 1000' or 'ping 12.112.7.173 data aaaa size 1000' and/or 'ping 12.112.7.173 data 0000 size 1000'). With that being said, your error rate isn't horrible (average of one out of 15876 packets), but certainly it can't hurt to have the telco test it, since its a new circuit. Most likely the telco will loop the circuit and not find any issues; sometimes the problems mysteriously disappear after the telco tests it Perhaps can be of further help? I'll try the pings. I don't know that I can do a loopback test being MPLS. It's not like a clear T where I control both sides.

One side is my CPE back to ATT, then the same at the other side. So it's ATT who would have to loop it.

Trying to explain what I wanted them to do there would just be too painful. I did manage to get ftp servers up and running on either side of the link. I'll do something like zip the entire Solarwinds toolset together ten times to create on big honking file and then pull it back and forth, maybe while also running Wankiller. That ought to fill that baby up with traffic. Pull that file back and forth twenty times and I should sniff out any trouble with the line.

The problem up here in the northeast is that facilities stink. I've spent more time over the years arguing with Verizon (now Fairpoint) about whether it's my equipment or their lines causing the problem. I don't know that I'll ever get this thing down to zero. Thought I would post a quick follow up. I did try the pings with the all 1's and all 0's patterns. Those didn't produce any problems. But I have this big, honkin' 679MB file I created by zipping a bunch of crap together.

I've found that I can't pull or push that thing across the link without generating somewhere between 100 and 300 input errors consistently. And that's the only set of machines using the link. Free Drum Fills Music Written.