Collapsed and lateral rings - info from ICG

Back to home page

From: Doquisa, Christina
Sent: Tuesday, August 16, 2005 11:13 AM
Subject: Collapsed and Latteral Rings - Transport Network


I'm working on a project that will identify all single point of failures across all network types. I found the information below interesting and beneficial for the whole team and wanted to forward it to you.

Christina Doquisa
Manager, Network Operations
Mpower Communications

* Change is the only constant *

From: Heley, Curt []
Sent: Friday, August 12, 2005 9:41 AM
To: DeRidder, Edward; Doquisa, Christina; Doquisa, Christina; Kramer, Paul
Cc: Denning, Mark
Subject: RE: Collapsed and Latterals

Just so we're on the same page, here's some definitions of the terms I think you're looking for:

Collapsed ring -- a protected system (with service and protect or G1 & G2 paths) where both paths are in the same cable, creating a single point of failure. Nearly all our systems are protected, although we do have some collapsed rings.

Lateral -- the path between the backbone or main loop cable and the building being served. We have a lot of laterals.

Single entrance -- when a building is served by a single lateral cable between the backbone cable and the POP in the building. The fibers are typically spliced into the backbone ring in different directions creating diversity at the backbone, but the path is collapsed between the backbone and the POP creating a single point of failure for that segment. This is an accepted practice for laterals shorter than a certain distance, in our case, 300-500 feet, depending on some circumstances. Most of our buildings are on-net this way.

Dual (also called "diverse") entrances -- when a building has two lateral cables between the backbone and the POP in the building creating true diversity between the POP and the serving switch/collector node. Typically COs and IXCs have dual entrances whenever feasible (Las Vegas has only one dual entrance CO, South South).

Backbone/main ring or loop/distribution loop -- the preferred architecture for metro networks. A high fiber count cable leaves the switch or collector node site and passes each target building in a given geographical area returning to the opposite side or diverse entrance of the switch or collector node site. Laterals to buildings are spliced into the backbone cable with some fibers pointing each direction, providing diversity around the loop back to the switch. Most of N Cal and San Diego were built this way, the LA-Orange County networks were not.

I think what you're asking is how many buildings are single threaded for a longer than acceptable distance, either because they were constructed that way or because they were incorrectly spliced at the backbone. There are several here in N Cal:

185 Berry, San Francisco (serves Level 3 and Sprint dark fiber) 1624 Franklin, Oakland (long single entrance serving Pac West, iParadigm and Qwest dark fiber, all but Qwest diverse at the backbone) 675 N 1st Street, San Jose (very long lateral, diverse at the backbone, serves Western States Teleport) 310 N 10th, Sacramento (very long lateral, diverse at the backbone, OC48)

There were more buildings that fit this category, but those are currently stubbed with no service. Also, there are diversity issues at various buildings which OSP is clearing up through maintenance and grooming projects, one example being the AT&T POP at 611 Folsom in San Fran. Diverse fibers exist, but splicing and a traffic roll is required to make that site truly diverse. This will be completed as time allows. Other sites may also exist, however, we are still in the process of auditing our networks (again, as time and funding allows) and will have to deal with additional sites as they are found.

In addition, there's a lot of unprotected circuits in the network, I don't track those, but one example that comes to mind is the OC48 from OnFiber between 770 L Street and 55 S Market. That system does not have a protect path.

Denning will have to get you his list when he gets back in next week.

Hope that helps! Please let me know if you have any questions.

Curtis R. Heley
Senior Network Engineer - Outside Plant - CA
Ph: (408) 380-0925
Fax: (408) 513-0907
Cell: (415) 720-3816

-----Original Message-----
From: DeRidder, Edward
Sent: Thursday, August 11, 2005 4:28 PM
To: Doquisa, Christina; Doquisa, Christina; Kramer, Paul
Cc: Heley, Curt; Denning, Mark
Subject: RE: Collapsed and Latterals


All of our major fiber rings are diversified in Northern California. I will be reviewing the Southern Calfornia systems with Paul next week, but any rings collapsed onto single fiber bundles would be best addressed by what ICG referred to as Outside Plant management (Curt Heley and Mark Denning)


Are you aware of any collapsed rings in the network currently? The last issue I was aware of was between 274 brannan St and 620 3rd St in San Francisco and the fiasco with Southern Cal Edison… but both issues have since been resolved. Any others you can think of?


-----Original Message-----
From: Doquisa, Christina []
Sent: Thursday, August 11, 2005 4:19 PM
To: Doquisa, Christina; DeRidder, Edward; Kramer, Paul
Subject: Collapsed and Latterals

Do we have any Latteral or Collapsed rings in our fiber network?

Christina Doquisa
Manager, Network Operations
Mpower Communications

* Change is the only constant *

Back to home page