At this point, all circuits have been clear of errors for the last 24 hours, with the exception of Adtran3, t1 0/4, which had 235 CSS/235 ES. The recommendations by jayh have reduced the errors dramatically, I'm marking as correct. I'm marking olivertwisted suggestion as helpful as well. Because of the external data T1, I cannot act on it, but it's a good idea, nonetheless, and would probably correct the remaining issue.