No problem - just a trench

[RumorConsumer]

 

 

Delight with me in my trench! Digging a line for our generator to power the walk in fridge in the event of a power outage. Sneaking in a comm conduit as well to change an important repeater to gateway. 12” separation between conduits and they’re at different depths and all that spec. Using CAT7 as I have some 10GbE capable NAS hardware around that we might want to use around campus not just in its little closet. And high well shielded bandwidth is just exciting. 

Ok one question. Why is 10GbE copper over RJ45 (And it’s later variants) so much more expensive than SFP+ kit? 

[NolanHerring]

Can't say for certain but here are some links that might provide more information:

https://community.fs.com/blog/should-we-use-cat6-or-cat6a-for-10gbase-t.html

https://www.datapro.net/techinfo/cat7_info.html

https://nascompares.com/2018/06/01/should-you-chose-10gbase-t-copper-over-sfp-for-10g-ethernet/

I've always been a fan of doing fiber outdoors myself as it 100% eliminates the risk of lightning/electrical issues etc.

[NolanHerring]

Also when you mentioned CAT7 my mind skipped a beat, because I legit didn't realize it was a thing. Turns out it kind of isn't =P

https://www.youtube.com/watch?v=9ZqzyXexE7Y

[RumorConsumer]

Ah well that is a bummer. All good though, my 200ft outdoor “CAT7” able to be buried cable was not too expensive. CAT6a speeds and shielding will do.

[PhilipDAth]

Nice trenching!  Are you dropping in just copper because its less than 100m?  Or are you putting in fibre as well?

>Ok one question. Why is 10GbE copper over RJ45 (And it’s later variants) so much more expensive than SFP+ kit? 

I haven't seen anyone use Cat7.

Cat6a is normally suitable for 10Gbe copper, and it should not be that expensive.  It is normally cheaper than fibre and 10Gbe SFP+ adaptors ... but in your case you would probably need to use externally rated copper in a special cladding (sometimes grease filled), and yep that costs more.

If you use fibre - you can either put it in conduit or buy it with an external cladding and then yoiu can drop it straight in.  You also don't require spacial seperation for fibre (although your might have local building codes or regulatory requirements for this).

With fibre you can start with Gigabit SFP adaptors, and then at some point in the future upgrade for 10Gbe SFP+.

You might want to think if it is worth going to 10Gbe straight away, or weather you are doing 10Gbe because you can and it would be a "nice to have".  After all, everything is currently running over a a WiFi link ...

I'd be tempted to drop in a 6 core fibre.  You could initially use a pair of them in a LACP channel.  You could use cheaper MS120-8(P)'s at each end with Gigabit SFPs.  The MS120-8's also come in a PoE option which could be usefull in your case.  Of course, you might already have an MS at each end.

If you jump all the way up to an MS225-24(P) which has 4 x 10 SFP+ ports.  You use 10Gbe fibre between them, and then plug in your NAS via 10Gbe SFP+ to the MS225.

[RumorConsumer]

@PhilipDAth I really do just heat the kiwi now.

 Yeah honestly it’s only 54m total distance and will be supplying a wired connection to a single MR52 serving a 100 person dining hall, kitchen, and weight room all in one building. It’s exceptionally rare that all those people would be there with devices let alone doing anything but eating. Maybe up to 15 laptops every now and then. And the NAS thing is less a likelihood to ever happen in this building. I decided no on fiber because copper just seemed easier and there is nothing that actually calls for that kind of firepower. I’m open to being flipped on it but for what i see this doing “CAT7”/6a seems to be just fine. Definitely no MS anywhere on our site, and in this case like I said it’s a nice to have and yes everything would be wireless.

Now the rest of our site is getting OM3 fiber for the buildings that actually need it both in capability and in distance.

[PhilipDAth]

Think about the lifetime you are expecting from this service going into the trench.  Do you expect it to last several decades without being touched?  Or maybe your site is dynamic and you know it probably only needs to last 10 years, or perhaps you know the site doesn't like to be touched so you need to get 30 years out of it.

>Now the rest of our site is getting OM3 fiber for the buildings that actually need it both in capability and in distance.

OM3 is fine for everything today.  Typically I get customers to install OM4 these days.  If it was me I would get a price on OM4 as well before making a final decision.  OM4 is definately more expensive.

If it was me, I would definately install more than one type of whatever circuit you run.  Lets say 5 years from now you are getting some weird circuit problems - having a spare cable lets you switch between them for testing.

Also their is nothing worse than when a cable fails (perhaps water ingress, perhaps a rodent, perhaps a tree root, etc) and you have to dig the whole lot up again.  Having a spare to change over makes it so much easier.

If there is enough money - I would drop a couple of fibres into the trench - and just don't bother terminating them (to save money).  Just leave them coiled up at each end.  Then if something changes into the future, they are ready to be used.  Fibre itself is usually not that expensive.  It's all the labour to fit them off, trench them, etc which makes up most of the cost.

I understand your logic with a run that is only 54m.  But I'm with @NolanHerring .  I'd be dropping a 6 core OM4 fibre into the trench.  The fibre will be good for decades to come.  It wont be affected by lightening strikes, temperature or water ingress, and is likely to be supported by new network standards for several decades.

[RumorConsumer]

@PhilipDAth @NolanHerring You guys are like the stern internet savvy fathers I never had 🤣

Well jeez you are all making a good case for fiber. I will call lanshack on Monday and price it out. We cannot afford to get unterminated cable and have somebody come out and fit them off to LC or TC or whatever and I dont want to learn to do that part myself as relatively simple as it looks. Getting it pre-terminated and then buying and installing my own wall mount termination box (https://www.lanshack.com/1-panel-Wall-Mount-Termination-Box-P6549.aspx) doesn't look that hard and I have professionally installed examples on property I can use as a reference. A little more money now. Im leaning! Im leaning!

[Uberseehandel]

Possibly because I have spent much time living on the Pacific Rim (earthquakes > flexible wooden buildings > high danger of fires) and consequent draconian building codes, I automatically think "fibre" when presented with connecting buildings (or high speed PtP/PtMP wireless links).

Differences in Electrical Potential are a fire risk. Interestingly, there was a thunderstorm the other day around Wellington NZ, the thunder was noticeably louder than residents had previously experienced. I am sure it was no coincidence that the atmosphere was full of ash and debris from the bush fires raging in Australia, creating conditions under which differences in potential develop. Damage was done, the automated conservatory roof venting and shade control system was completely burnt out. The replacement is going to be opto-isolated, it could happen again at any time.

I'm in the process of bringing my fibre notes up to date.

[PhilipDAth]

>Differences in Electrical Potential are a fire risk

Interesting you should mention that.  I had a business client in a high rise building (they had many floors).  The building had internal trunking that ran around the outside edge of the building.  Consequently a lot of the Ethernet cabling was run around the outsde in this trunking.

They experienced repeated serious equipment damage.  My personal guess was the wind was causing static build up on the outside skin of the building and this was being induced into the Ethernet cabling killing the kit.

I also feel it increased the risk of lightening damage - but they had kit being damaged so frequently it could not have been lightening alone.

Eventually they did a refit and the cabling used ducts dropped from the roof instead of the outside edge.

Problem didn't happen anymore.

[SoCalRacer]

My recommendation would be fiber.

My thought why 10Gb RJ45 Copper is more expensive than fiber is that it is new. Fiber has been around for 30+ years where 10Gb RJ45 Copper much less.

[RumorConsumer]

That makes sense. I think we will do the fiber option. I cannot imagine this area needing more speed than what OM3 has to offer for a very long time and I will compare costs. I definitely want to future proof without going nuts. More than that, this building has some very old electrical wiring that I believe I have had issues with in the past and do not care to repeat that in any way with this new home run. I can see running the copper and then experiencing weirdness and having the question of whether the old power lines are in some way affecting things. No thank you. So fiber it will be. I have some apprehension about running fiber but from what I have read its actually less delicate than copper and as long as you pull using the kevlar part that is meant to take several hundred pounds of pressure its all gravy. We are only going a short distance as well so I think we'll be ok. 

[RumorConsumer]

Progress

 

[RumorConsumer]

Aaaand we're done - except some cable management. 

Works super well. Really happy with how it turned out. 

Always, always, always use cable lube. Life changer. 

@PhilipDAth @NolanHerring @SoCalRacer @BrechtSchamp