Apple TV and ethernet switch

The purpose is to correct confusing, misleading, or inaccurate statements, which are not helpful to anyone.

Your assertion that a hub should be used instead of a switch seemed without factual basis. Since that was germane to the OP's situation, it was challenged. Your subsequent, somewhat self-contradictory explanation that you meant a "switching hub" seemed to me to undermine a pair of (AFAIK) accurate, detailed answers to one of the OP's questions ("Is the ethernet switch the same as a hub?"). My reply addressed that. I tried to avoid as much as possible inflammatory statements without compromising a clear explanation of the facts as I understand them, which I believe is substantially helpful to the OP.

Substituting a hub for a switch will not remedy his problem or improve network throughput. If we can agree on that much, I think we should move on.

Actually, in reviewing the originator's initial post, he was asking if a switch is the same as a hub. It is and it isn't. It is certainly not the case that the names "hub" & "switch" are well established throughout the world.

I have spent a great deal of time over the past 10 years dealing with network issues in our locations outside of the U.S. in countries where the indigenous language is not English. A switch is a switching hub, pure and simple. It has come to be referred to simply as a switch in the U.S. and perhaps Britain but I can tell you that in Asia it is still frequently referred to as a hub. An Ethernet bridge is also referred to as a switch in some circles. This is certainly not an area where terminology has been clarified and agreed upon by all.

Depending on the network, there isn't necessarily a clear, measurable advantage in having a switching hub over a plain old-fashioned hub. For the average home network I doubt that it would make much of a noticeable difference either way. When you're dealing with this issue at a corporate level, however, there is a significant difference in cost between the two.

I may have fumbled a hastily typed response, but this is an area I know quite well. I call it a picayune point of disagreement because I don't believe the particular type of hub, or switch, if you need that term for clarity, is going to be a significant factor in anyone's Apple TV problems.

Whatever missteps, confusion, disputes or disagreements that may arise from any given person's post, I think a greater measure of civility will go a lot further in achieving clarity.

iMac Mac OS X (10.4.9) The more I think, the more I think I shouldn't think more.

I was indeed trying to clear up any confusion about the differences between hubs & switches, but I don't believe it was a "picayune" point in the context of this topic. While I agree that I also sometimes benefit from that kind of discussion, I think we are obligated to weigh that benefit against the need to remain on topic.

I'll admit I sometimes have trouble with that. It is particularly difficult to decide what to say when attempting to clear up a confusing or contextually misleading statement would take the discussion far off its original topic. If one is brief & to the point, it is often interpreted as nit picking or a personal affront; being too verbose runs the risk of promoting threadjacking. Saying nothing is safest but clears up nothing.

I think all regular posters face this dilemma from time to time. Hopefully, we can be as tolerant of others in this respect as we would like them to be of ourselves.

It is more than semantics, a hub is not a switch, and a switch is not a hub. The only commonality is that they allow you to connect multiple devices to your lan. A switch is the way to go, but a good hub will likely work...

With all due respect, regardless of local slang or custom or whether you call it a "switching hub" or simply a "switch," the device the label refers to has capabilities a standard hub (or concentrator, or whatever you wish to call it) lacks, & should be the preferred choice in essentially every network a home or small office user would construct.

This based both on the technical explanation Winston mentioned & my own practical experience with networks as small as a few devices. Like you, I find it a bit surprising that it makes a noticeable difference on small networks, but it does seem to, at least during large file transfers.

How so? What will be the noticeable difference in the performance of an Apple TV between the two?

iMac Mac OS X (10.4.9) The more I think, the more I think I shouldn't think more.

a hub is not a switch, and a switch is not a hub

Of course, that's true technically, but a switch is also referred to as a switching hub.

but a good hub will likely work...

That has been precisely my point. After all of this rancor, I don't think there is a performance difference that is significant enough to have a noticeable effect on a typical home network.

iMac Mac OS X (10.4.9) The more I think, the more I think I shouldn't think more.

I have been reading/watching the exchanges thus-far, and I feel obligated to now explain why, in MY home network, I went from a Hub to a Switch and the performance gains that I have noticed. YMMV.

First, let's agree that a "Hub" operates at the Physical Layer (Layer 1) of the 5 Layer (TCP/IP) or the 7 Layer (ISO/OSI) model. And, let's also agree that a "Switch" operates at Layer 2 (Data Link Layer) of these models.

Now, my network in brief:
2 Vista PCs, wired gigabit
2 Macs, wired gigabit
1 Mac, wireless (g)
3 Apple TVs, two wired, one wireless (n)
1 Yamaha MusiCast system, wireless to 3 MusiCast Clients, wireless
1 Yamaha Receiver (plays net radio), wired 100
2 AirPort Express Base Stations, wireless (g) for music
1 Phaser Printer, wired 100
3 DirecTV HR20s (recording satellite boxes with the ability to play local music), wired 100

For me, upgrading to a Switch yielded an order of magnitude (10x) improvement in syncing my Apple TVs, with a quiescent network. With an "active" network (file transfers in parallel to Apple TV syncing/copying new media), no performance degradation as compared to the quiescent network.

Is my home network "typical"; probably not statistically. Is it 'outrageously different' from some others that have posted here? No.

Of course, if the 'network' consists of an Internet connection, 1 computer and 1 Apple TV, you will not be generating enough packets to measure the performance improvements of a Layer 1 device as compared to a Layer 2 device. Conversely, add devices, and the benefits become clearly apparent in favor of a Layer 2 device.

Cheers,

Thanks for providing your own information as an example. I wouldn't consider your home network to be a typical one, though. You obviously have a data intensive multimedia network running, so I can see where a switch might provide a noticeable improvement in sync time for your three Apple TVs. With such a data intensive network even before you got your Apple TVs a couple of months ago, however, I'm curious as to why you weren't already using a switch.

I wasn't using a Layer 2 Switch due mainly to frugality & lack of attention to my "infrastructure" (i.e., Buying new devices and not paying attention to my existing connections).

I had a 16 port Netgear 10/100 Layer 1 Hub, but after many years of faithful service, it finally gave up the ghost shortly after I added the Apple TVs. Jealousy no doubt on the part of the Hub 🙂

Before the Hub went to the great silicon heaven in the sky, I did have data collected via SNMP and wall-clock timings for transferring files to/from various devices on my LAN, so I re-ran some of the transfers with the Layer 2 Switch in place; I shouldn't have been surprised by the performance increase, but I was.

Cheers,

I shouldn't have been surprised by the performance increase, but I was.

Also in line with the point I was trying to make. You actually measured the performance difference. Before that, your system was functioning sufficiently such that you didn't feel the need to upgrade (at least until you got your Apple TVs). From the standpoint of frugality, a more modest network may not see as significant a gain in performance.

In any case, for your network, a switch is obviously the best way to go.

iMac Mac OS X (10.4.9) The more I think, the more I think I shouldn't think more.

How so? What will be the noticeable difference in
the performance of an Apple TV between the
two?

I can't say, never having tried that particular comparison. However, I can say that I noticed a distinct improvement in file transfer times between two iMacs whenever there was a lot of bidirectional network traffic between them. It wasn't as dramatic as 10X or anything like that, but it was enough to convince me that switching is a cheap, simple method of improving network performance, even in tiny networks. As the network grows, so do the improvements, sometimes dramatically. This seems reason enough to recommend switches over hubs.

This is really all (or more than) I feel comfortable saying about this in this thread. If you want to discuss it further, please start a topic focused on this particular issue.