Re: MCI WorldCom fiber cut - Syracuse, NY
Could the explanation be simpler? Effects of gas pipeline and water main breaks tend to be localized because they supply commodity goods, and there is local storage (and, especially in case of water, local supplies) of those. Hence such breaks affect fewer people. The gas supply to my kitchen does not depend on maintenance of uniform pressure in all the gas pipelines from the well off the shore of Louisiana all the way to New Jersey; my supplier has enough gas stored around here to keep pumping for quite a while even if a pipeline in Kentucky is cut. On the other hand, when a fiber gets cut in Ohio, and I am trying to get some bits from California, it does not help me to know that somebody in Pennsylvania has terabits on her server that she is willing to ship to me. Andrew At 07:27 PM 10/5/99 -0400, Mike Heller wrote:
I find it interesting that we don't hear about more accidental gas explosions and water main breaks from backhoes. Are the blueprints and posted warnings more detailed or is there more fiber than utility pipe in the ground?
Michael Heller Sr. Systems Engineer Earthweb, Inc. 212.448.4175 mikeh@earthweb.com
followed by other contributions from Derek Balling, Charles Sprickman, and Nathan Stratton. ************************************************************************ Andrew Odlyzko amo@research.att.com AT&T Labs - Research voice: 973-360-8410 http://www.research.att.com/~amo fax: 973-360-8178 ************************************************************************
Andrew, by far one of the best explanations I've read! Excellent illustrations. You're suppositions are indeed correct. In addition to which, fiber doesn't emit a nice electrical signature that can be detected easily, making it hard to avoid. Plastic, glass, fiberglass, kevlar and the other elements of most fiber runs lay invisible to many detection devices that rely upon metals content or electrical impulse emission (crosstalk, noise, EMF...) for detection purposes. Now, some have written that we should encase these things with various high-strength metals. I'm not willing, as an end consumer, to bear the increased overall costs being passed to me, because $VBC laid 10,000 miles (16 000 km) of protectively-encased fiber. Costs would be staggering. In addition, repairs and splices more difficult in those situations where a backhoe manages to ding up one of these things and cause an actual cut. In my part of Ohio, the engineering maps get updated way to infrequently to suit my comfort level. We have a 3 year old fiber run into our NOC that is still not known to most of the high-cap techs that come out here. In fact, the local gas company was boring new pipe into the ground a few months ago, and weren't even aware of the fiber laying 50 feet (17 m) away. All other services we're clearly marked. As already stated, since fiber doesn't kill, people are complacent and wreckless around it. -Matt. Andrew Odlyzko wrote a while back: < < < Could the explanation be simpler? Effects of gas pipeline and < water main breaks tend to be localized because they supply < commodity goods, and there is local storage (and, especially in < case of water, local supplies) of those. Hence such breaks < affect fewer people. The gas supply to my kitchen does not < depend on maintenance of uniform pressure in all the gas pipelines < from the well off the shore of Louisiana all the way to New Jersey; < my supplier has enough gas stored around here to keep pumping for < quite a while even if a pipeline in Kentucky is cut. On the other < hand, when a fiber gets cut in Ohio, and I am trying to get some < bits from California, it does not help me to know that somebody < in Pennsylvania has terabits on her server that she is willing < to ship to me. < < Andrew < -- Matthew D. Lammers, Columbus, Ohio, US
"Matthew D. Lammers" wrote:
Andrew, by far one of the best explanations I've read! Excellent illustrations. You're suppositions are indeed correct.
In addition to which, fiber doesn't emit a nice electrical signature that can be detected easily, making it hard to avoid. Plastic, glass, fiberglass, kevlar and the other elements of most fiber runs lay invisible to many detection devices that rely upon metals content or electrical impulse emission (crosstalk, noise, EMF...) for detection purposes.
When gas installers lay underground plastic gas lines (I have such running to my backup generator) they lay a wire in the ground next to the pipe. The wire is not attached to anything at the ends, just wrapped around the gas line where it comes out of the soil. The idea is to use this with a signal generator/receiver, much like the devices we use to chase down wire pairs. Addition of an electrical conductor, with attachment points for tracers might make a LOT of sense. The next thing done when an underground gas pipe is laid, is a mylar film is buried, a foot or so above the pipe, which has all kinds of warnings on it about the gas line below. The tape is mirror-reflective mylar. It's sure to catch some attention if the backhoe operator doesn't wail right through and beyond... Perhaps those laying fiber should consider (perhaps they already do?) such methods. It's unclear if there are laws against putting up "gas pipeline, do not dig" signs where there really aren't gas pipelines, but that might be another way to deter the backhoes... Finally, though, I wonder if any of the fiber cable manufacturers have data on the susceptibility of their jacket material to the various gases pumped by the gas pipeline companies. Laying the fibers inside active gas pipelines would sure be a deterrent to those digging, though in the case of a cut, you'd need a significant piece of fiber to replace that part which had melted in the fire... -- ----------------------------------------------------------------- Daniel Senie dts@senie.com Amaranth Networks Inc. http://www.amaranthnetworks.com
Let's not forget backhoe-operator school. Probably about 6 hours are dedicated to safety around power lines, gas lines, etc. I don't think "other utility" services and even mentioned. Deepak Jain AiNET On Wed, 6 Oct 1999, Matthew D. Lammers wrote:
Andrew, by far one of the best explanations I've read! Excellent illustrations. You're suppositions are indeed correct.
In addition to which, fiber doesn't emit a nice electrical signature that can be detected easily, making it hard to avoid. Plastic, glass, fiberglass, kevlar and the other elements of most fiber runs lay invisible to many detection devices that rely upon metals content or electrical impulse emission (crosstalk, noise, EMF...) for detection purposes.
Now, some have written that we should encase these things with various high-strength metals. I'm not willing, as an end consumer, to bear the increased overall costs being passed to me, because $VBC laid 10,000 miles (16 000 km) of protectively-encased fiber. Costs would be staggering. In addition, repairs and splices more difficult in those situations where a backhoe manages to ding up one of these things and cause an actual cut.
In my part of Ohio, the engineering maps get updated way to infrequently to suit my comfort level. We have a 3 year old fiber run into our NOC that is still not known to most of the high-cap techs that come out here. In fact, the local gas company was boring new pipe into the ground a few months ago, and weren't even aware of the fiber laying 50 feet (17 m) away. All other services we're clearly marked. As already stated, since fiber doesn't kill, people are complacent and wreckless around it.
-Matt.
Andrew Odlyzko wrote a while back: < < < Could the explanation be simpler? Effects of gas pipeline and < water main breaks tend to be localized because they supply < commodity goods, and there is local storage (and, especially in < case of water, local supplies) of those. Hence such breaks < affect fewer people. The gas supply to my kitchen does not < depend on maintenance of uniform pressure in all the gas pipelines < from the well off the shore of Louisiana all the way to New Jersey; < my supplier has enough gas stored around here to keep pumping for < quite a while even if a pipeline in Kentucky is cut. On the other < hand, when a fiber gets cut in Ohio, and I am trying to get some < bits from California, it does not help me to know that somebody < in Pennsylvania has terabits on her server that she is willing < to ship to me. < < Andrew <
-- Matthew D. Lammers, Columbus, Ohio, US
In addition to which, fiber doesn't emit a nice electrical signature that can be detected easily, making it hard to avoid. Plastic, glass, fiberglass, kevlar and the other elements of most fiber runs lay invisible to many detection devices that rely upon metals content or electrical impulse emission (crosstalk, noise, EMF...) for detection purposes.
Now, some have written that we should encase these things with various high-strength metals. I'm not willing, as an end consumer, to bear the increased overall costs being passed to me, because $VBC laid 10,000 miles (16 000 km) of protectively-encased fiber. Costs would be staggering. In
You wouldn't need to encase it. Bury a little bit of copper with it, and blast RF out of it (think of it is a locater service). --msa
I was under the impression that fiber trunks used to be buried (circa 15 years ago) with a copper tracer in them. Then there was some good reason why they were no longer done that way. Like corrosion or something. Deepak Jain AiNET On Wed, 6 Oct 1999, Majdi Abbas wrote:
In addition to which, fiber doesn't emit a nice electrical signature that can be detected easily, making it hard to avoid. Plastic, glass, fiberglass, kevlar and the other elements of most fiber runs lay invisible to many detection devices that rely upon metals content or electrical impulse emission (crosstalk, noise, EMF...) for detection purposes.
Now, some have written that we should encase these things with various high-strength metals. I'm not willing, as an end consumer, to bear the increased overall costs being passed to me, because $VBC laid 10,000 miles (16 000 km) of protectively-encased fiber. Costs would be staggering. In
You wouldn't need to encase it. Bury a little bit of copper with it, and blast RF out of it (think of it is a locater service).
--msa
deepak@ai.net said:
I was under the impression that fiber trunks used to be buried (circa 15 years ago) with a copper tracer in them. Then there was some good reason why they were no longer done that way. Like corrosion or something.
</CYNIC> cost <CYNIC> Without wishing to blow my own country's trumpet, one of the few things right about the UK (and to a great extent Europe's) telecommunications carrier market is sensible telcos dig proper ducts, put sensible fiber in them, bury them at sensible depths, and in general only provision SDH (read SONET in the US). And *seem* to keep maps. Even US carriers in the UK who commonly have US fiber cuts do this. -- Alex Bligh GX Networks (formerly Xara Networks)
Without wishing to blow my own country's trumpet, one of the few things right about the UK (and to a great extent Europe's) telecommunications carrier market is sensible telcos dig proper ducts, put sensible fiber in them, bury them at sensible depths, and in general only provision SDH (read SONET in the US). And *seem* to keep maps.
We have a very comprehensive map of the ducting that is laid. And another layer that covers fibre which extends into buildings, which is a very impressive system. You can look at a building like Telehouse and find the fibre paths very quickly. It can plot routes based upon specified optical power budgets and redundancy requirements, and generate a works list for joins etc if a new ring is being built. Regards, Neil. -- Neil J. McRae - Alive and Kicking. neil@DOMINO.ORG
Speaking of fiber cuts, there's an MCI/Worldcom fiber run in downtown Houston, TX that's out. The location is at Walker and Louisiana as best I can tell, which just so happens to be in front of the building my office is in. There's a long line of MCI Worldcom trucks and a crew out there now. Several multi-homed customers of UUNet in Houston have their UUNet links down. I'll send more info as I have it. Both Qwest and BBN/GTEi have POP's in this building, but they appear to be unaffected. UUNet's main Houston POP is a few hundred feet down Louisiana. With all the construction in this area of downtown, I'm frankly shocked that this hasn't happend until now. -- Joseph W. Shaw - jshaw@insync.net Freelance Computer Security Consultant and Perl Programmer Free UNIX advocate - "I hack, therefore I am."
It would also be a lot of fun to see all the gory details about how exactly things got cut, and what they had to do to fix it. Maybe it was easy, maybe they ran out of slack in the cable and needed to wait hours for more cable to arrive, maybe the one guy who knows how to fix things is on vacation, etc. I spoke to someone at MCI WorldCom Fiber Security who claims they have an organized repository for stuff like this on their intranet. If you're a customer, they said the only way to find out about current outages is to call the MFS/Brooks/whatever trouble number you have, or try to get your sales rep to get post-mortem info if your outage was long enough to trigger SLA credits. -- Scott M. Drassinower scottd@cloud9.net Cloud 9 Consulting, Inc. White Plains, NY +1 914 696-4000 http://www.cloud9.net On Wed, 6 Oct 1999, Joe Shaw wrote:
Speaking of fiber cuts, there's an MCI/Worldcom fiber run in downtown Houston, TX that's out. The location is at Walker and Louisiana as best I can tell, which just so happens to be in front of the building my office is in. There's a long line of MCI Worldcom trucks and a crew out there now. Several multi-homed customers of UUNet in Houston have their UUNet links down. I'll send more info as I have it.
Both Qwest and BBN/GTEi have POP's in this building, but they appear to be unaffected. UUNet's main Houston POP is a few hundred feet down Louisiana.
With all the construction in this area of downtown, I'm frankly shocked that this hasn't happend until now.
-- Joseph W. Shaw - jshaw@insync.net Freelance Computer Security Consultant and Perl Programmer Free UNIX advocate - "I hack, therefore I am."
participants (9)
-
Alex Bligh -
Andrew Odlyzko -
Daniel Senie -
Deepak Jain -
Joe Shaw -
Majdi Abbas -
Matthew D. Lammers -
Neil J. McRae -
Scott Drassinower