Question Are Switches Basically Plug n Play?

[Caveman]

I'm pretty sure the switch in my system that feeds the house is non functional. I tried to unplug/plug to reset and no dice... Every connection in the house that's tied to the switch is non functional, so I have to assume a bad switch. Is there any other thing I should do before buying a new switch?

Also, can anyone recommend a good, simple, 16 port reliable gigabit ethernet switch??

 

[SamirD]

A 100Mb switch, or any Ethernet device should have Auto MDIX. The MDI/MDIX switch or button was needed on 10Mb devices, so switches could be daisy-chained. Of course there were some low end 100Mb stuff that lacked Auto MDIX.

That is incorrect. I have a switch we purchased back when 100Mbps was like 10Gbps today and both of those 100Mbps uplinks are manual MDIX. I thought the switches were bad until I realized I needed to manually set the duplex on the nic that I was testing with to get proper speeds. They better still work since they were $900/ea back then!

 

[SamirD]

TBH.... 2GB WAN probably isn't worth it to you.
Not unless your doing pretty large hosting / FTP server / Remote backups.
Also you will need to get new gear across everything.
Its not just because you have 2gb WAN you automatically get on it.
Your modem will most likely need to be replaced.
Your network cards unless your on the new Z470 / X570 / B550 (the last 2 can vary) will all need to get upgraded.
Then there is the switch. Consumer end, unless your rolling from scratch with all new gear which has it built in, IMO its not really worth it at the level your probably sitting at.

And at the cost of moving to 2.5Gbps, 10Gb starts to make more sense...

 

[SamirD]

The patent for Auto MDIX was published 2001. I remember devices older than that. (They had "lifetime warranty" ... and died.)

I have functioning devices older than that.

 

[ch33zw1z]

Yea, the 2.5 thing is so BS IMO, 10gbps should be produced for the lower end market at this point. It's ridiculous to craft 802 standards for in between speeds. If you can build a 2.5/5, you can build a 10. It's just more money grabbin

 

[MtnMan]

That is incorrect. I have a switch we purchased back when 100Mbps was like 10Gbps today and both of those 100Mbps uplinks are manual MDIX. I thought the switches were bad until I realized I needed to manually set the duplex on the nic that I was testing with to get proper speeds. They better still work since they were $900/ea back then!

It was typically the low end stuff that didn't have Auto MDIX, but that's not an exclusive club. Truthfully "most" 100Mb stuff is going to have Auto, including NICs.

 

[MtnMan]

I have functioning devices older than that.

In use?

Some of the first networks I worked on were 10Base2 and 10Base5, and Token Ring. I wired the house with thinnet coax in probably 1990.

 

[SamirD]

Yea, the 2.5 thing is so BS IMO, 10gbps should be produced for the lower end market at this point. It's ridiculous to craft 802 standards for in between speeds. If you can build a 2.5/5, you can build a 10. It's just more money grabbin

But issue is the cabling. It's the same issue when the industry moved from 10Mb to 100Mb--cabling kept a lot of people from upgrading. It's why the 100VG-AnyLAN standard was such a strong competitor to the 100Mb that became mainstream. To be honest, I don't know why 100VG didn't win because the nics were only a bit more and would have saved massive amounts of money in plant re-cabling as well as had other advantages:

100BaseVG - Wikipedia

en.wikipedia.org

 

[SamirD]

It was typically the low end stuff that didn't have Auto MDIX, but that's not an exclusive club. Truthfully "most" 100Mb stuff is going to have Auto, including NICs.

Ummm...this wasn't low end at all--it was one of the first switches that came with 2x 100Mbps uplinks that were full-duplex. There were many years where 100Mbps was a standard and Auto MDIX wasn't around.

 

[SamirD]

In use?

Some of the first networks I worked on were 10Base2 and 10Base5, and Token Ring. I wired the house with thinnet coax in probably 1990.

They're pluged in and working, but I only use them when I want to see what happens when I run a speed test on a 10Mbps port.

Wow, 10Base5--I have heard about how bad that was. I worked with 10Base2 when buidling my first network--it's length limits were handy for sure.

 

[ch33zw1z]

But issue is the cabling. It's the same issue when the industry moved from 10Mb to 100Mb--cabling kept a lot of people from upgrading. It's why the 100VG-AnyLAN standard was such a strong competitor to the 100Mb that became mainstream. To be honest, I don't know why 100VG didn't win because the nics were only a bit more and would have saved massive amounts of money in plant re-cabling as well as had other advantages:

100BaseVG - Wikipedia

en.wikipedia.org

I've been working in data centers of all kinds for 10+ years, I have never once encountered any 2.5BaseT installs, cards, switches, cabling, etc...

I work outside DC's as well, never seen it there.

1gbps is plenty for most users, faster wireless speeds are coming, and most dc gear is 10gbps and up already.

My XP, my .02 is that it's going to keep 10gbps prices inflated longer.

 

[MtnMan]

They're pluged in and working, but I only use them when I want to see what happens when I run a speed test on a 10Mbps port.

Wow, 10Base5--I have heard about how bad that was. I worked with 10Base2 when buidling my first network--it's length limits were handy for sure.

10Base5 was OK, if you had some good vampire taps, and left them alone. Only had a couple of segments of it, but it was a solution to span 500 meters between buildings.

 

[SamirD]

I've been working in data centers of all kinds for 10+ years, I have never once encountered any 2.5BaseT installs, cards, switches, cabling, etc...

I work outside DC's as well, never seen it there.

1gbps is plenty for most users, faster wireless speeds are coming, and most dc gear is 10gbps and up already.

My XP, my .02 is that it's going to keep 10gbps prices inflated longer.

Yeah, that doesn't surprise me since it's basically the same spec for 1GBase-T. It's actually too new for anyone to be doing anything serious with it. And with use 10Gb prices where they are, unless someone is trying to run something faster over existing wiring, 10Gb and DACs bring a lot of value.

 

[SamirD]

10Base5 was OK, if you had some good vampire taps, and left them alone. Only had a couple of segments of it, but it was a solution to span 500 meters between buildings.

Yeah, that's basically what I read about them is that as long as they were installed and left alone, all was good. That's some amazing distances too.

 

[VirtualLarry]

lol! They should have just added some LEDS and called them 'gamer class' switches and sold a ton more, haha.

LOL. Why not? Asus does that.

 

[VirtualLarry]

Yea, the 2.5 thing is so BS IMO, 10gbps should be produced for the lower end market at this point. It's ridiculous to craft 802 standards for in between speeds. If you can build a 2.5/5, you can build a 10. It's just more money grabbin

You volunteering to upgrade all of the businesses' building wiring that used cat5e, for free? Get crackin'.

 

[VirtualLarry]

My XP, my .02 is that it's going to keep 10gbps prices inflated longer.

that's what the vendors want, too. They don't want us mere peons to have $5 10gbE-T NICs built into our motherboards, they want to be able to charge us $100-200 extra for that privilege.

 

[ch33zw1z]

You volunteering to upgrade all of the businesses' building wiring that used cat5e, for free? Get crackin'.

Nothing's free bruh. Those switches aren't cheap, and it's only 8 ports each.

 

[ch33zw1z]

that's what the vendors want, too. They don't want us mere peons to have $5 10gbE-T NICs built into our motherboards, they want to be able to charge us $100-200 extra for that privilege.

You seem content with that. As you said, are you offering to go upgrade all those businesses building wiring for free? I think not, so enjoy getting less for the same price. I'll be content to point out how it's a day late and a dollar short. Wireless is still right on it's heels, wpa3 coming too.

 

[VirtualLarry]

As you said, are you offering to go upgrade all those businesses building wiring for free? I think not, so enjoy getting less for the same price. I'll be content to point out how it's a day late and a dollar short.

Again, and no hard feelings, but if you were a business-owner, would you prefer:

1) ripping out all of the wiring, just to get the next network "speed boost", or
2) take advantage of newer standards, that allow you to change out the switches and NICs, and get a 2.5x-5x boost in network "performance", without changing the wiring?

If those standards didn't exist, as you seem to want them not to exist, but if there were ONLY the choice between being stuck at 1Gbit/sec over Cat5e, versus having to completely rip+replace, even if the 10GbE-T gear was cheap(er), then there's still significant labor and cost for wiring upgrades.

And I'm sorry, but wireless is never going to be more than a niche, utilized by "mobile" devices like laptops and phones. It will NEVER be a real replacement for wired networking, for fixed workstations and the like.

 

[ch33zw1z]

Again, and no hard feelings, but if you were a business-owner, would you prefer:

1) ripping out all of the wiring, just to get the next network "speed boost", or
2) take advantage of newer standards, that allow you to change out the switches and NICs, and get a 2.5x-5x boost in network "performance", without changing the wiring?

You seem to be under the impression that's the only two options.

3. Not bother because 1gbps is just fine. Option is to not spend any money at all, because 1gbps or wifi ac is plenty for almost everyone.

If those standards didn't exist, as you seem to want them not to exist, but if there were ONLY the choice between being stuck at 1Gbit/sec over Cat5e, versus having to completely rip+replace, even if the 10GbE-T gear was cheap(er), then there's still significant labor and cost for wiring upgrades.

Again, this wasn't even a thought until 8 years ago, a full 10 years after 10gbps was a standard. That was 20 years ago in 2002, only 3 years after 1gbps was standardized.

And I'm sorry, but wireless is never going to be more than a niche, utilized by "mobile" devices like laptops and phones. It will NEVER be a real replacement for wired networking, for fixed workstations and the like.

We can agree to disagree here. It's used extensively in the field, by millions of workstations right now, both laptops and desktops.

I'm not sure what world you live in, but laptops and phones are a huge piece of the tech market including in the Corp world. I literally repair laptops for mom+pop shops up thru big corps, they're hardly niche and neither is wifi, users are on wifi like crazy.

If ieee rolled out the standards 10 years ahead of when they did, I wouldn't being saying anything, but at this stage of the game, it's just a money grab.

The 10gbps market will remain inflated, as you even agreed with in post 41. You can pay to upgrade, twice, on your way there.

10gbps between the switch and nas will be great, but nothing else, even 4k video, will saturate the gig bus or even the wifi link to an end point.

Meanwhile, DC's are pushing 40 and 100 gig links now.

Edit: I should add, the most likely deployment of "multigigabit" that I can foresee enterprise-wise is to power wifi 802.11ax WAP's that are coming, since they would tx/rx enough data to saturate a 1gig link. Not a single user, but many on the AP will.

 

[VirtualLarry]

Edit: I should add, the most likely deployment of "multigigabit" that I can foresee enterprise-wise is to power wifi 802.11ax WAP's that are coming, since they would tx/rx enough data to saturate a 1gig link. Not a single user, but many on the AP will.

This is true. Also, consumer DOCSIS 3.0/3.1 modems support modulation rates in excess of 1Gbit/sec in throughput (in the aggregate). Comcast is actually advertising their "Gig" as 1.2gbit/sec, and should be provisioned as 1.440Gbit/sec.