LAN to LAN (Switch to Switch) Max ethernet length 1 gig (1000tx)

[fihunts]

While I have read that there is a 328 feet or 100 meter max when connecting from one switch to a device (PoE).  However, my question is this.  If I have a Cat 6 or Cat 6a direct connect ethernet cable that runs 372' from one switch to another switch and I have constant traffic coming and going from both switches.  Am I still limited to the rules of 328' feet.  Since technically, each switch if I am correct and not doing any POE devices should have 328' reach?  Should I not have a max reach of 656' feet?  Not saying I want to try that.  But, right now I am running a 372' switch to switch and not having any drops, connection issues or latency.

Next question since I am not having any issues.  What might be the possible downfalls going beyond the "so called" 328' max? other than the 500 nanosecond delay (latency) for every 3 feet beyond 328' which to me is nominal or not even going to bother me.

I am not a schooled network engineer, I am a hands on database guy, but I dabble in networking.  So, In layman's terms.  Can someone tell me what my possible pitfalls are if any?

[marce1000]

 - 

The maximum distance for Ethernet cable runs is typically limited by the signal attenuation and crosstalk caused by the cable itself, and not by any limit imposed by the switch. The distance limits for Ethernet cables are typically around 100 meters (328 feet) for standard Cat5e or Cat6 cables. However, newer cables such as Cat6a and Cat7 are designed to reduce attenuation and crosstalk, and therefore can support longer cable runs.

In your case, you mention that you are using a Cat6 or Cat6a cable that runs for 372 feet between two switches and you are not experiencing any issues such as drops, connection issues, or latency. This suggests that your cable is working well within its specifications, and you are not experiencing any signal degradation that would cause problems.

It's worth noting that the 328 feet (100 meter) max distance is the standard distance limit set by IEEE 802.3ab. If you are using a switch that supports POE (Power over Ethernet) it should be noted that the distance limit for a POE device will be less than 100 meters.

As for the possible downfalls of going beyond the 328 feet max, the main concern would be signal attenuation and crosstalk caused by the cable, which can result in errors and reduced signal quality. Additionally, as you mention, there may be a slight increase in latency for signals traveling greater distances, but this is likely to be negligible for most applications.

          - So in general for business networking , only use supported lengths only for the per cable category , 

 M.

[Joseph W. Doherty]

"What might be the possible downfalls going beyond the "so called" 328' max?"

The main one, as already mentioned a couple of times in @marce1000's posting, is attenuation.   I.e. the signal gets weaker as distance increases, until the far side cannot reliable decode information (see SNR).

Have you ever used walkie-talkies? The have range limitations based upon distance.  Close, they work great.  Far, they don't "work" at all.  Borderline between close and far, they sort of work, i.e. you can understand part of a transmission, or, possibly all of it, with "say again".

Likewise with wired signals too, copper or fiber.

Last place I worked, we had one much over 100m Ethernet cable.  Ran it at 10 Mbps (on Cat5?).  Lots and lots of errors (about 50% or so, corrupted frames, I recall) across link, but with retransmissions, information was passed back and forth.

Also, BTW, Ethernet standards specific what bandwidth rates should work at 100m (for that category specification).  However, some cables are better than minimum spec.  I.e. you can run them even further than 100m, but you've "voided" your performance guarantee.

[fihunts]

I have done some reading and for the most part.  I guess the IEEE and the groups that have put together these standards wrote them for those who wish to maximize the full capabilities of Cat 6/6a and earlier version of Cat cables.  "The so-called 100 meters refers to that if over 100meters, it cannot pass the related technical indicators in 1000M bandwidth, and brings some problems such as a decrease in speed."  "https://www.linkedin.com/pulse/how-does-data-transmission-network-cable-100-meters-max-gary-han/"

Which is basically what you guys have said and other things that I should be aware of.  So, by going over the max directed by the IEEE.  Be prepared for potential performance issues.  Having good cabling at the length I have gone over, "approx 13%" Does not seem to have impacted my network.

Thanks guys!

[Leo Laohoo]

---

@fihunts wrote:
Next question since I am not having any issues.  What might be the possible downfalls going beyond the "so called" 328' max? other than the 500 nanosecond delay (latency) for every 3 feet beyond 328' which to me is nominal or not even going to bother me.

---

For non-PoE port(s), the speed drops down to 10 Mbps, half-duplex.

For PoE port(s), PoE cannot get that far so the equipment manufacture will look for someone to blame why their PTZ camera keeps crashing.  And because I'm the network guy who answered the call, it is "guilty until proven otherwise", even though they were the ones who organized for a cable run that went >110 metres.  

[fihunts]

Hehe, I am well aware of the restrictions with the PoE.  I would never open a ticket on a POE device full knowing that I went over the 100 meter limitation.  My questions were specifically about going from a Cisco 10G Catalyst 6800 to a Cisco FPR 2100 non PoE.  I think between you three guys I have gotten what I needed to know.  Sounds like I am not gonna be in too much trouble going over by 12 meters.

Thanks again! 

[Leo Laohoo]

I would never envisage using copper cable for 10 Gbps.  I always use optics because it is more reliable than copper.

[fihunts]

We don't use that 10G Catalyst for 10 gb.  We use it for selling 1 gb bandwidth or less.  The Catalyst itself doesn't have any fiber slots on it.. I might need to double check, but the whole switch is rj45's.. It doesn't have any SFP slots to go fiber or ethernet.  I think it might be pretty old.  It works really well from what I can tell though.  I will need to get back to you on that though.

[fihunts]

As a matter a fact I know it is all rj45's.  That is all that is plugged into it.

[fihunts]

I won't know till tomorrow.. but I think this is what it is...  

Cisco Catalyst 6800 Series 10 Gigabit Ethernet Fiber Module with dual DFC4X

Visit

$30,656.99*

The C6800-32P10G-XL Ethernet module provides thirty two 1G or thirty two 10G transceiver ports.

[Joseph W. Doherty]

Oh my, Leo, you're a fiber (or fibre, down under?) bigot!  ; )

Leo is correct if he has in mind many of fiber's actual advantages over copper, but sometimes when you start to tote up the cost of using fiber ports and their transceivers, copper alternatives, financially, can start to become very attractive.

I suspect Leo might not be always against using copper for 10g, e.g. between a TOR switch and servers with 10g (copper) ports.

Again, though, there are advantages to fiber.  One interesting article I read, years ago, made a good case for SM fiber to even the desktop.

[Leo Laohoo]

LOL

I have been to a site, the first I've ever seen, where each desk had two pairs of MM SC connections.