Pat Sync and G.821 Statistics 1

[mikepiper]

New to the forum here, and hoping there's a few wizards that can help me. I'm struggling in all the googling I've done to come up with my answers. We have many different DS1's that we hand-off to Centurylink for different types of voice traffic. In monitoring some trouble circuits with our Sunrise Telecom Sunset T3 meters, I've noticed groups of DS1's that come across with a Pat Sync light and G.821 statistics, and other groups that don't. Same line coding and all (AMI/SF), but a push of the re-sync button will do nothing on one set, and will establish the Pat Sync on the others.

Can anyone help me digest this? From everything I've read about G.821, it was established as a threshold for quality of data transmission on a T1. Does this mean that our DS1's that are establishing Pat Sync are of a lesser quality than those that don't, or am I way off? Any help would be much appreciated.

Thanks,

Mike

 

[Dexman]

Are any errors such as timing slips or framing being detected by either the CPE or Centurylink on the DS1 level? Any DS3 errors being detected?

Has Centurylink intrusively tested the DS1 circuits in question? If so, did the test results show any problems?

I "FEATURE 00

 

[mikepiper]

I would think our switch would be what you'd consider the CPE in this case, and no, Its not picking up anything. I came in this morning and from across the room i could see that both meters had taken errors overnight, but were clear now. The DS1 with no pat sync established had taken 15 UAS and 12 OOF overnight on the DS1 level. The other circuit (which was flashing red/green on errors and pat sync) had absolutely clean looking DS1 measurements, not a single Unavailable or Errored Second. But, since pat sync was established, its counting these G.821 statistics and there were 433 UAS during around a 16 hour time frame.

When its taking these G.821 UAS, the Meter is saying "No Sync" up in the top right for Rcv:. It usually says "1-8 INV", and i have no idea what that means. The circuit that does not establish pattern sync always says "Rcv: Live"

 

[libellis]

The “1-8 INV” probably means a standard “1-in-8” test pattern, and the “INV” likely means it is inverted (an option on test equipment), so that the 8-bit pattern consists of seven “ones” and one “zero”. From your post it seems as though you only want to monitor links and aren’t setup with a loop back (or paired up with another meter) at the far end for testing via generated patterns, so not sure what’s happening. Is the Carrier generating a test pattern, or is your meter inadvertently set to inject one?

I don’t think the Pat Sync light would be on when monitoring a live circuit with random data. It would be on, for example, when a link is being tested with a pattern generated by the meter and with a loop back at the far end.

 

[mikepiper]

Reading through the User Manual of the Sunset T3 today, I kind of come to the same conclusion as far as thinking its something that had to do with test patterns, and shouldn't be on a LIVE circuit. I'm glad you clarified.

There should be no test patterns being generated by us or Centurylink on this circuit (under normal circumstances), but Centurylink does have horrible outside plant after our meet point before they get to their CO. Every single time we've had troubles, its been on their plant. Thinking back there were times that their CO tech was blaming us and telling us she was getting all 1111's from us (we were not generating anything, they replaced a faulty repeater and the circuit came up clean). I'm wondering if there is a chance that more faulty Centurylink plant is inducing these test patterns on this DS1. If thats the case, wouldn't it effect the whole DS1 and we would see ES, SES or UAS at the DS1 level?

 

[libellis]

A fault wouldn't likely mimic a test pattern, but when there is a Red Alarm condition (OOF threshold exceeded), and the Yellow Alarm comes on, a 16-bit (8 ones and and 8 zeros) pattern is repetitively sent. I suppose - depending upon the Carrier's monitoring and action policy - a loop back and test mode might be automatically activated at some point, but I don't really know.