37 errors is NOT a lot in a 24 hour period. My experience is that if there is some major problem you'll have about 20,000 errors in the first 2 seconds.

When I tested Microwave circuits we used to be allowed 1 error in 24 hours. Otherwise we had to tear the circuit down and try it again. We once did a pair of T-3 backup circuits for Bell Atlantic that ran through a repeater site across the Hudson before they got to their final destination. We could get that one down to about 5 errors in a 24 hour period and they were thrilled.

Microwave of course could take errors if a bird flew in the path of the beam. I don't think you're having that problem (the circuit isn't partially wireless, is it?), but 37 errors while not good is probably not even noticeable. Remember a T-1 is 1.544 million bits per second. That's about 92 million bits per minute, 5.5 billion bits per hour and approximately 133 billion bits per day. (I think I've got the math right.) So you've got 37 bad bits and 133 billion good ones.

You want to get it fixed, but not to agonize.

Sam


"Where are we going and why are we in this hand basket?"