Hi NANOG, Could someone please help me with a fast ethernet problem I am having. We have a POP in a 27 floor building, and have a rj45 run from the the bottom of the building (in the telco room) to the top of the building. We have cisco switches on either end and we are seeing about 5-20% packet loss on the trunk. Are we running into a distance limitation of fast ethernet, or are we suffering from another problem? I read that 328 feet is the limitation of fast ethernet. Is there anything short of getting a repeater for the cable run that I can do to boost the signal? TIA for your help. Ciao. Steve Rude
100 meters is, in fact, the distance limitation for Fast Ethernet, but you can usually exceed that if the link is full duplex. Note that I'm not recommending that you do so, just stating that it is possible. If your run length is more than 100 meters, and you're running half duplex, then I would say that is definitely your problem. -- Bruce Robertson, President/CEO +1-775-348-7299 Great Basin Internet Services, Inc. fax: +1-775-348-9412 http://www.greatbasin.net
And you are using shielded cable, correct? Best regards, ______________________________ Al Rowland
-----Original Message----- From: owner-nanog@merit.edu [mailto:owner-nanog@merit.edu] On Behalf Of Bruce Robertson Sent: Friday, January 10, 2003 12:19 PM To: Steve Rude Cc: nanog@merit.edu Subject: Re: fast ethernet limits
100 meters is, in fact, the distance limitation for Fast Ethernet, but you can usually exceed that if the link is full duplex. Note that I'm not recommending that you do so, just stating that it is possible.
If your run length is more than 100 meters, and you're running half duplex, then I would say that is definitely your problem.
-- Bruce Robertson, President/CEO +1-775-348-7299 Great Basin Internet Services, Inc. fax: +1-775-348-9412 http://www.greatbasin.net
On Fri, 10 Jan 2003, Al Rowland wrote:
And you are using shielded cable, correct?
Nah, I'm guessing he strung bare copper seperated by cotton balls. That's what I like to use in my 27-floor 100tx runs. Andy xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Andy Dills 301-682-9972 Xecunet, LLC www.xecu.net xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Dialup * Webhosting * E-Commerce * High-Speed Access
Works fine if you do it using x.25 :-) On Fri, 10 Jan 2003, Andy Dills wrote:
On Fri, 10 Jan 2003, Al Rowland wrote:
And you are using shielded cable, correct?
Nah, I'm guessing he strung bare copper seperated by cotton balls. That's what I like to use in my 27-floor 100tx runs.
Andy
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Andy Dills 301-682-9972 Xecunet, LLC www.xecu.net xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Dialup * Webhosting * E-Commerce * High-Speed Access
Actually andy, the oc192 wiccs in the 2600 series work better. :) On Fri, 10 Jan 2003, Andy Dills wrote:
On Fri, 10 Jan 2003, Al Rowland wrote:
And you are using shielded cable, correct?
Nah, I'm guessing he strung bare copper seperated by cotton balls. That's what I like to use in my 27-floor 100tx runs.
Andy
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Andy Dills 301-682-9972 Xecunet, LLC www.xecu.net xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Dialup * Webhosting * E-Commerce * High-Speed Access
putting a shield on cat5 or 6 cable doesn't significantly increase the noise rejection vs utp cat 5 at 100mb/s, you're shielding already balanced cable pairs. moreover they're signifcantly harder to install since they need to be properly grounded and shielded at both ends. joelja On Fri, 10 Jan 2003, Andy Dills wrote:
On Fri, 10 Jan 2003, Al Rowland wrote:
And you are using shielded cable, correct?
Nah, I'm guessing he strung bare copper seperated by cotton balls. That's what I like to use in my 27-floor 100tx runs.
Andy
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Andy Dills 301-682-9972 Xecunet, LLC www.xecu.net xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Dialup * Webhosting * E-Commerce * High-Speed Access
-- -------------------------------------------------------------------------- Joel Jaeggli Academic User Services joelja@darkwing.uoregon.edu -- PGP Key Fingerprint: 1DE9 8FCA 51FB 4195 B42A 9C32 A30D 121E -- In Dr. Johnson's famous dictionary patriotism is defined as the last resort of the scoundrel. With all due respect to an enlightened but inferior lexicographer I beg to submit that it is the first. -- Ambrose Bierce, "The Devil's Dictionary"
-----Original Message----- From: joelja@darkwing.uoregon.edu [mailto:joelja@darkwing.uoregon.edu] Sent: Friday, January 10, 2003 13:26 To: andy@xecu.net Cc: alan_r1@corp.earthlink.net; steve@rudedogg.com; nanog@merit.edu Subject: RE: fast ethernet limits
<snip>
moreover they're signifcantly harder to install since they need to be properly grounded and shielded at both ends.
I've seen people use shielded CAT5 to protect it from interference but they didn't bother grounding the shielding on either end....
On Fri, 10 Jan 2003, Stephen Fisher wrote:
I've seen people use shielded CAT5 to protect it from interference but they didn't bother grounding the shielding on either end....
In the "me too" category, I've seen a company install wireless on top of the Netherland's highest building (The Rembrandt's tower), which included using a lightning arrestor, and not ground it. It only took 3 storms (and 3 wireless cards) before they finally decided to ground the thing, and at the time, they weren't that cheap either :) Paul -- God devised pigeons as a means of punishment for man. Probably after the destruction of Sodom and Gomorrha he wanted to make sure that people would never again feel comfortable enough in a city to repeat the sins committed there, and he created the pigeons as a means to make the city dwellers' lives more miserable, as a constant reminder of their past sins.
AGREED, one end and one end only, or youre asking for a ground loop....ground the end with the best, shortest path to earth ground.....in his case, that would prob be the telco room end, "usually" theres a decent ground there somewhere..... Mileage may differ... At 16:30 1/10/03 -0600, you wrote:
Joel Jaeggli wrote: [...]
moreover they're signifcantly harder to install since they need to be properly grounded and shielded at both ends.
I've actually seen some very impressive ground loops. I'd ground one end. (Actually I'd use fiber, but hey.)
Peter E. Fry
On Fri, 10 Jan 2003 17:53:11 EST, blitz said:
AGREED, one end and one end only, or youre asking for a ground loop....ground the end with the best, shortest path to earth ground.....in his case, that would prob be the telco room end, "usually" theres a decent ground there somewhere.....
Anybody else ever chased a grounding wire you suspected wasn't doing its job, only to find that it was quite securely clamped to a large drain pipe only 4 feet from where it left the building? Sounds good, except that the pipe transitioned from cast iron to PVC 2 feet from the exit point....
Yes, seen that and worse.. How about the di-electric fittings on either a water heater, or a water meter? Of course, "no one" would use a HOT water line as a ground, but it does happen, and I believe the NEC requires both be bypassed now with a conductor the same size as power neutral. I routinely jump unbridged meters and water heaters (cold pipe to hot) with a #6 when in doubt. Just to be sure. I've got lots of #6 in the scrap box, and plenty clamps... Cheap insurance..
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 - --On Friday, January 10, 2003 17:53:11 -0500 blitz <blitz@macronet.net> wrote:
AGREED, one end and one end only, or youre asking for a ground loop....ground the end with the best, shortest path to earth ground.....in his case, that would prob be the telco room end, "usually" theres a decent ground there somewhere..... Mileage may differ...
Some hours reading the back issues of the journal found at http://www.compliance-club.com will hopefully inform you why star grounding is a thing of the past. Ground both ends. If you are afraid of ground loops, place a heavy (as in 10-16mm2 or AWG way below 10) ground conductor alongside the signal cable, and ground it firmly in both ends. That will take the current away from the shields. Ungrounded shileds are inefficient for EMI and RF shielding, while at times efficient AC hum blockers. And, IANAEE, but I've played with big sound systems that exhibit all these problems. - -- Måns Nilsson Systems Specialist +46 70 681 7204 KTHNOC MN1334-RIPE We're sysadmins. To us, data is a protocol-overhead. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.7 (OpenBSD) iD8DBQE+Ibr902/pMZDM1cURAt+XAKCFc+MVCS5WfkfJGCk5yO+iSVNtEQCfR3Ri 85ebf8wkeypAWWkuRaHbnis= =GiNg -----END PGP SIGNATURE-----
Date: Sun, 12 Jan 2003 19:59:08 +0100 From: =?ISO-8859-1?Q?M=E5ns_Nilsson?= <mansaxel@sunet.se> Sender: owner-nanog@merit.edu
Some hours reading the back issues of the journal found at http://www.compliance-club.com will hopefully inform you why star grounding is a thing of the past. Ground both ends. If you are afraid of ground loops, place a heavy (as in 10-16mm2 or AWG way below 10) ground conductor alongside the signal cable, and ground it firmly in both ends. That will take the current away from the shields.
Ungrounded shileds are inefficient for EMI and RF shielding, while at times efficient AC hum blockers.
And, IANAEE, but I've played with big sound systems that exhibit all these problems.
I find this simply frightening! Have you any idea how much potential of ground can vary in a large building? It's easy to have AMPS of current flowing through the shield of a cable and enough voltage offset to be dangerous to people. (802.3 allows enough breakdown potential that the equipment is unlikely to have a problem, though.) Grounded at one end is better than floating for shielding, so this almost reasonable (as long as the wiring is all installed to spec) but grounding at both ends in the wrong environment can lead to serious problems. Since the 802.3 sections on 10Base-T does not deal with shielded wire at all, there is nothing there on the subject. But other sections on 10Base5 and 10Base2 are explicit that 10Base5 MUST be grounded at exactly one point and 10Base2 recommends that one point be grounded when the cable extends out of a room. More than one ground is explicitly prohibited. The archives of comp.dcom.lans.ethernet are full of people who have high error rates because of multiple grounds. R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: oberman@es.net Phone: +1 510 486-8634
On Mon 13 Jan 2003 (08:33 -0800), Kevin Oberman wrote:
Date: Sun, 12 Jan 2003 19:59:08 +0100 From: =?ISO-8859-1?Q?M=E5ns_Nilsson?= <mansaxel@sunet.se> Sender: owner-nanog@merit.edu
Some hours reading the back issues of the journal found at http://www.compliance-club.com will hopefully inform you why star grounding is a thing of the past. Ground both ends. If you are afraid of ground loops, place a heavy (as in 10-16mm2 or AWG way below 10) ground conductor alongside the signal cable, and ground it firmly in both ends. That will take the current away from the shields.
Ungrounded shileds are inefficient for EMI and RF shielding, while at times efficient AC hum blockers.
And, IANAEE, but I've played with big sound systems that exhibit all these problems.
I find this simply frightening! Have you any idea how much potential of ground can vary in a large building? It's easy to have AMPS of current flowing through the shield of a cable and enough voltage offset to be dangerous to people. (802.3 allows enough breakdown potential that the equipment is unlikely to have a problem, though.)
Grounded at one end is better than floating for shielding, so this almost reasonable (as long as the wiring is all installed to spec) but grounding at both ends in the wrong environment can lead to serious problems.
Since the 802.3 sections on 10Base-T does not deal with shielded wire at all, there is nothing there on the subject. But other sections on 10Base5 and 10Base2 are explicit that 10Base5 MUST be grounded at exactly one point and 10Base2 recommends that one point be grounded when the cable extends out of a room. More than one ground is explicitly prohibited.
The archives of comp.dcom.lans.ethernet are full of people who have high error rates because of multiple grounds.
Do also consider that a one end grounded shield in these circumstances is a health and possibly fire hazard. If you are running cable in a building where ground has significant potential differences between the two ends of the cable, you have a serious shock hazard occurring in a place where people aren't likely to expect or take precautions against it. I don't know what the current carrying capacity of a shield is, so how much of a fire risk you might run if your one end only sheild were to become two ended. My answer would be that if this situation exists, either get it fixed or use optical fibre. -- Jim Segrave jes@nl.demon.net
I find the same Kevin..I've done a lot of work in broadcast stations as well, and ground loops are a constant problem. Hum is introduced into audio lines, even in balanced pairs, and Cat5 is not much different. In a high rise, I can see a neutral failing somewhere on a high floor, and that piece of #10 going incandescent, setting fire to anything combustible between the floor its terminated on and earth ground. (The resistance of an old steel framed building is NOT always lower than that piece of copper, especially old riveted buildings). Maybe they do it differently in the EU, but fire safety is also a concern to me. Thank you, but I'll ground nearest to the earth, common bonded point. Its served me well over the years.
Some hours reading the back issues of the journal found at http://www.compliance-club.com will hopefully inform you why star grounding is a thing of the past. Ground both ends. If you are afraid of ground loops, place a heavy (as in 10-16mm2 or AWG way below 10) ground conductor alongside the signal cable, and ground it firmly in both ends. That will take the current away from the shields.
Ungrounded shileds are inefficient for EMI and RF shielding, while at times efficient AC hum blockers.
And, IANAEE, but I've played with big sound systems that exhibit all these problems.
I find this simply frightening! Have you any idea how much potential of ground can vary in a large building? It's easy to have AMPS of current flowing through the shield of a cable and enough voltage offset to be dangerous to people. (802.3 allows enough breakdown potential that the equipment is unlikely to have a problem, though.)
Grounded at one end is better than floating for shielding, so this almost reasonable (as long as the wiring is all installed to spec) but grounding at both ends in the wrong environment can lead to serious problems.
Since the 802.3 sections on 10Base-T does not deal with shielded wire at all, there is nothing there on the subject. But other sections on 10Base5 and 10Base2 are explicit that 10Base5 MUST be grounded at exactly one point and 10Base2 recommends that one point be grounded when the cable extends out of a room. More than one ground is explicitly prohibited.
The archives of comp.dcom.lans.ethernet are full of people who have high error rates because of multiple grounds.
R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: oberman@es.net Phone: +1 510 486-8634
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 - --On Monday, January 13, 2003 13:35:14 -0500 blitz <blitz@macronet.net> wrote:
I find the same Kevin..I've done a lot of work in broadcast stations as well, and ground loops are a constant problem. Hum is introduced into audio lines, even in balanced pairs, and Cat5 is not much different.
Read the articles at www.compliance-club.com -- there is a two-piece article on audio systems, by the technical manager at Cadac. It definitely goes against all gut reactions we've forced ourselves into; especially in the telco/audio business. But it makes sense! It works! And, usually the audio folks were the last to find out: Look at your average network component; the grounding scheme of a big modern router. It is mesh grounds all over the place -- and because it is the only way it will ever get the box approved by FCC/UL/CSA or any similar agency.
In a high rise, I can see a neutral failing somewhere on a high floor, and that piece of #10 going incandescent, setting fire to anything combustible between the floor its terminated on and earth ground. (The resistance of an old steel framed building is NOT always lower than that piece of copper, especially old riveted buildings).
Which is why one must construct a low-resistance/impedance mesh, connecting shields and grounds all over the place, so as to short out the potentials. The steel frame could serve as one of these grids, but probably needs to be augmented. And, I wrote "AWG way below 10"... If it is not enough, get a thicker one, and get a new electrical contractor. (see below)
Maybe they do it differently in the EU, but fire safety is also a concern to me.
If you have potential differences like these the house needs a new electrical grid. Wait, that means that the entire North American continent needs rewiring ;-) Anyway, the mains PE MUST be made the best path home, and in a fail-safe manner.
Thank you, but I'll ground nearest to the earth, common bonded point. Its served me well over the years.
Look at the emissions that manage to get into a star grounded, one point only system vs emissions bounced off a mesh grounded system with a decent EMC testing rig and repeat that phrase afterwards ;-) - -- Måns Nilsson Systems Specialist +46 70 681 7204 KTHNOC MN1334-RIPE We're sysadmins. To us, data is a protocol-overhead. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.7 (OpenBSD) iD8DBQE+I1NT02/pMZDM1cURAtDHAJ9AwED4nokCNsotQjrNzbkG751m5wCfT2/R a5z6Hij53cuK4UZTwdHxqiE= =elgF -----END PGP SIGNATURE-----
Maybe they do it differently in the EU, but fire safety is also a concern to me.
If you have potential differences like these the house needs a new electrical grid. Wait, that means that the entire North American continent needs rewiring ;-) Anyway, the mains PE MUST be made the best path home, and in a fail-safe manner.
Thank you, but I'll ground nearest to the earth, common bonded point. Its served me well over the years.
Look at the emissions that manage to get into a star grounded, one point only system vs emissions bounced off a mesh grounded system with a decent EMC testing rig and repeat that phrase afterwards ;-)
I thought earthing one side of a shielded cable (be it Cat5, or any other type) was actually part of electrical regulations. (maybe I just assumed that as a result of practice) It is a known fact that earth varies within buildings, and from place to place in the ground, this is nothing to do with faulty wiring in houses or fault north american continent, this is a simple chemical/physical phenomenon as a result of subtle changes in water tables, salt, minerals etc So this means there is a difference in potential from point to point - thats voltage to you and me. So if you connect both of these together you get a flow of current - not good.. Of course, when you ground one end it also means you should be careful when working at the other end not to touch the shied as you are earthed to local ground and it may therefore have a live voltage should you touch it. Steve
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 - --On Tuesday, January 14, 2003 12:31:24 +0000 "Stephen J. Wilcox" <steve@telecomplete.co.uk> wrote:
I thought earthing one side of a shielded cable (be it Cat5, or any other type) was actually part of electrical regulations. (maybe I just assumed that as a result of practice)
It varies with local regulations. In .uk, I'd say you need to ground the screen, because a floating screen with potential difference is considered a hazard at work.
It is a known fact that earth varies within buildings, and from place to place in the ground, this is nothing to do with faulty wiring in houses or fault north american continent, this is a simple chemical/physical phenomenon as a result of subtle changes in water tables, salt, minerals etc
So this means there is a difference in potential from point to point - thats voltage to you and me. So if you connect both of these together you get a flow of current - not good..
Current flow is inevitable. The key is to shunt it away ASAP, where "S" means both "soon" and "safe", and to top those requirements, these connections need to look like attractive paths to RF energy, in order to cope with EM/RF interference rules.
Of course, when you ground one end it also means you should be careful when working at the other end not to touch the shied as you are earthed to local ground and it may therefore have a live voltage should you touch it.
See above. Summary: Ground often, ground well, mesh it, do away with long isolated runs, and in general, avoid big potentials. Shunt them. This starts to look like the "peer locally" argument. Good to be back in topic ;-) - -- Måns Nilsson Systems Specialist +46 70 681 7204 KTHNOC MN1334-RIPE We're sysadmins. To us, data is a protocol-overhead. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.7 (OpenBSD) iD8DBQE+JKmW02/pMZDM1cURAnLtAKCaLOfEDcqR1/pYjhBxONvj5UEZTgCeL+PM qa3HlMPD+shfM10Bq9mvHDE= =xD2q -----END PGP SIGNATURE-----
Dang. Snapple -> out nose. I hear aluminum coated dental floss is making a comeback in the wiring racket...
Nah, I'm guessing he strung bare copper seperated by cotton balls. That's what I like to use in my 27-floor 100tx runs.
Andy
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Andy Dills 301-682-9972 Xecunet, LLC www.xecu.net xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Dialup * Webhosting * E-Commerce * High-Speed Access
Steve, What type medium are you using? If it is normal Cat5/6 then the limitation is 100 meters for total distance and as you approach that limit the signal degrades. That said, 100baseFX can run for 400 meters due to the fact that it is fiber, both are part of the fast Ethernet specification though. A repeater would boost signal, but perhaps a switch in there might not be a bad idea, segment the 27 floors into VLANs, reduce overall traffic traveling between floors and eliminate the 27 floor run. Hope this helps, Kristian P. Jackson -----Original Message----- From: owner-nanog@merit.edu [mailto:owner-nanog@merit.edu] On Behalf Of Steve Rude Sent: Friday, January 10, 2003 3:01 PM To: nanog@merit.edu Subject: fast ethernet limits Hi NANOG, Could someone please help me with a fast ethernet problem I am having. We have a POP in a 27 floor building, and have a rj45 run from the the bottom of the building (in the telco room) to the top of the building. We have cisco switches on either end and we are seeing about 5-20% packet loss on the trunk. Are we running into a distance limitation of fast ethernet, or are we suffering from another problem? I read that 328 feet is the limitation of fast ethernet. Is there anything short of getting a repeater for the cable run that I can do to boost the signal? TIA for your help. Ciao. Steve Rude
Steve, What type medium are you using? If it is normal Cat5/6 then the limitation is 100 meters for total distance and as you approach that limit the signal degrades. That said, 100baseFX can run for 400 meters due to the fact that it is fiber, both are part of the fast Ethernet specification though.
And just for the record, 100BaseFX in full-duplex mode can actually run for 2000 meters with multimode fiber 62.5 micron core, 125 micron outer. -- Jorge Hernandez UNAM Network Operations Center http://www.noc.unam.mx (+52) 55 5622 8509
I used to see these exact same results when I would setup Wireless pop's on towers taller than 400Ft. I was able to push the envelope a bit, however when I saw the issues that you speak of, it was when I had bad crimps, or sometimes a bad cable all together. Cat5 should be fine for this... if you figure 12ft risers you are probably cutting it close on the distance but not going over it. -Scotty ----- Original Message ----- From: "Steve Rude" <steve@rudedogg.com> To: <nanog@merit.edu> Sent: Friday, January 10, 2003 3:01 PM Subject: fast ethernet limits
Hi NANOG,
Could someone please help me with a fast ethernet problem I am having. We have a POP in a 27 floor building, and have a rj45 run from the the bottom of the building (in the telco room) to the top of the building. We have cisco switches on either end and we are seeing about 5-20% packet loss on the trunk.
Are we running into a distance limitation of fast ethernet, or are we
suffering
from another problem? I read that 328 feet is the limitation of fast ethernet. Is there anything short of getting a repeater for the cable run that I can do to boost the signal?
TIA for your help.
Ciao.
Steve Rude
Previously, Steve Rude (steve@rudedogg.com) wrote:
Hi NANOG,
Could someone please help me with a fast ethernet problem I am having. We have a POP in a 27 floor building, and have a rj45 run from the the bottom of the building (in the telco room) to the top of the building. We have cisco switches on either end and we are seeing about 5-20% packet loss on the trunk.
Are we running into a distance limitation of fast ethernet, or are we suffering from another problem? I read that 328 feet is the limitation of fast ethernet. Is there anything short of getting a repeater for the cable run that I can do to boost the signal?
Well, when I don't have drawings or prints, I usually figure on 12 or 13ft per floor. So, figure somewhere between 324 and 351 feet from 1st floor to the ceiling on 27. For a run like this, you probably wanted to use fiber with media converters on each end. (Assuming you're not running switches with 100BaseFX ports or a GBIC slot...) -- Douglas A. Dever doug@e-xpedient.com 216.373.8517 - DID 216.401.5888 - Cell
On Fri, 10 Jan 2003, Douglas A. Dever wrote:
Previously, Steve Rude (steve@rudedogg.com) wrote:
Hi NANOG,
Could someone please help me with a fast ethernet problem I am having. We have a POP in a 27 floor building, and have a rj45 run from the the bottom of the building (in the telco room) to the top of the building. We have cisco switches on either end and we are seeing about 5-20% packet loss on the trunk.
Are we running into a distance limitation of fast ethernet, or are we suffering from another problem? I read that 328 feet is the limitation of fast ethernet. Is there anything short of getting a repeater for the cable run that I can do to boost the signal?
Well, when I don't have drawings or prints, I usually figure on 12 or 13ft per floor. So, figure somewhere between 324 and 351 feet from 1st floor to the ceiling on 27. For a run like this, you probably wanted to use fiber with media converters on each end. (Assuming you're not running switches with 100BaseFX ports or a GBIC slot...)
Indeed altho depending on purpose bear in mind a repeater is just a hub or switch, if you can fit one half way up you should solve your loss problems. Yeah 328 is the alleged limit but it can go further usually, as another reader said you need good connectors and cables tho. A more likely cause of trouble for you in the ducting is interference if you're running alongside AC mains and other data cables both of which will increase noise and reduce the limit even more. Steve
You could use fiber and a fiber conversion box. Or you could use a switch or repeater half way. On Fri, 10 Jan 2003, Steve Rude wrote:
Hi NANOG,
Could someone please help me with a fast ethernet problem I am having. We have a POP in a 27 floor building, and have a rj45 run from the the bottom of the building (in the telco room) to the top of the building. We have cisco switches on either end and we are seeing about 5-20% packet loss on the trunk.
Are we running into a distance limitation of fast ethernet, or are we suffering from another problem? I read that 328 feet is the limitation of fast ethernet. Is there anything short of getting a repeater for the cable run that I can do to boost the signal?
TIA for your help.
Ciao.
Steve Rude
you need to put a fluke lanmeter or similar device (with tdr) to validate the cable... you may just need to reterminate the ends, but it's also likely that it's simply way out of spec. joelja On Fri, 10 Jan 2003, Steve Rude wrote:
Hi NANOG,
Could someone please help me with a fast ethernet problem I am having. We have a POP in a 27 floor building, and have a rj45 run from the the bottom of the building (in the telco room) to the top of the building. We have cisco switches on either end and we are seeing about 5-20% packet loss on the trunk.
Are we running into a distance limitation of fast ethernet, or are we suffering from another problem? I read that 328 feet is the limitation of fast ethernet. Is there anything short of getting a repeater for the cable run that I can do to boost the signal?
TIA for your help.
Ciao.
Steve Rude
-- -------------------------------------------------------------------------- Joel Jaeggli Academic User Services joelja@darkwing.uoregon.edu -- PGP Key Fingerprint: 1DE9 8FCA 51FB 4195 B42A 9C32 A30D 121E -- In Dr. Johnson's famous dictionary patriotism is defined as the last resort of the scoundrel. With all due respect to an enlightened but inferior lexicographer I beg to submit that it is the first. -- Ambrose Bierce, "The Devil's Dictionary"
just go mm fiber.. Bri On Fri, 10 Jan 2003, Joel Jaeggli wrote:
you need to put a fluke lanmeter or similar device (with tdr) to validate the cable... you may just need to reterminate the ends, but it's also likely that it's simply way out of spec.
joelja
On Fri, 10 Jan 2003, Steve Rude wrote:
Hi NANOG,
Could someone please help me with a fast ethernet problem I am having. We have a POP in a 27 floor building, and have a rj45 run from the the bottom of the building (in the telco room) to the top of the building. We have cisco switches on either end and we are seeing about 5-20% packet loss on the trunk.
Are we running into a distance limitation of fast ethernet, or are we suffering from another problem? I read that 328 feet is the limitation of fast ethernet. Is there anything short of getting a repeater for the cable run that I can do to boost the signal?
TIA for your help.
Ciao.
Steve Rude
-- -------------------------------------------------------------------------- Joel Jaeggli Academic User Services joelja@darkwing.uoregon.edu -- PGP Key Fingerprint: 1DE9 8FCA 51FB 4195 B42A 9C32 A30D 121E -- In Dr. Johnson's famous dictionary patriotism is defined as the last resort of the scoundrel. With all due respect to an enlightened but inferior lexicographer I beg to submit that it is the first. -- Ambrose Bierce, "The Devil's Dictionary"
I believe your pushing the limits as to ethernet over Cat5. I can suggest you use the very best cable (shielded of course) you can get, and be meticulous in your connector installations and you might get away with it. Avoid other wiring if possible (fat chance huh?) and anything electrical interference producing, like fluoro ballasts, transformers etc. etc. Ground the end closest to earth ground to a good common point ground in the building, (not a power box) and leave the other end free floating and not touching anything electrically. 100 meters is supposedly the limit for ethernet, and "assuming" a 12' floor, your'e around 24 feet over spec. You might try to find some cat 6 cable if you can, its supposedly super premium cat 5, with better freq response and jittter control. http://www.controlcable.com/products/category_6_cable.html One last advice, use REAL good patch cables as well...they may help squeeze the last bit of performance out.. At 12:01 1/10/03 -0800, you wrote:
Hi NANOG,
Could someone please help me with a fast ethernet problem I am having. We have a POP in a 27 floor building, and have a rj45 run from the the bottom of the building (in the telco room) to the top of the building. We have cisco switches on either end and we are seeing about 5-20% packet loss on the trunk.
Are we running into a distance limitation of fast ethernet, or are we suffering from another problem? I read that 328 feet is the limitation of fast ethernet. Is there anything short of getting a repeater for the cable run that I can do to boost the signal?
TIA for your help.
Ciao.
Steve Rude
On Fri, 10 Jan 2003, blitz wrote:
100 meters is supposedly the limit for ethernet, and "assuming" a 12' floor, your'e around 24 feet over spec.
100 meters is primarily a "timing limit" for half duplex, ie if you're running half duplex, how large can your broadcast domain be before you risk losing small packets due to the collission not reaching all speakers on the wire (and this counting the max number of repeaters in a broadcast domain). 100/half will do padding of small packets to achieve just this effect. I've seen 170meter CAT5 100 meg full duplex links working just fine. Of course, as stated before, it's an interference thing, if you have a lot of noise around the wiring you'll shorten the usable length. Of course, this is a ethernet chipset issue, but there are definately chipsets out there capable of electronically driving links much more than 100 meters just fine. -- Mikael Abrahamsson email: swmike@swm.pp.se
participants (23)
-
Al Rowland -
Andy Dills -
blitz -
Brian -
Bruce Robertson -
Douglas A. Dever -
ed＠the7thbeer.com -
Jim Segrave -
Joel Jaeggli -
Jorge Hernandez -
K. Scott Bethke -
Kevin Oberman -
Kristian P. Jackson -
Mikael Abrahamsson -
Måns Nilsson -
Paul Wouters -
Peter E. Fry -
Robert A. Hayden -
Scott Granados -
Stephen Fisher -
Stephen J. Wilcox -
Steve Rude -
Valdis.Kletnieks＠vt.edu