Ask Slashdot: Is There Space For Open Hardware In Networking? 121
New submitter beda writes: Open hardware has got much attention with the advent of Raspberry Pi, Arduino and their respective clones. But most of the devices are focused either on tinkerers (Arduino) or most notably multimedia (Raspberry Pi). However, there is not much happening in other areas such as home routers where openness might help improve security and drive progress. Our company (non-profit) is trying to change this with Turris Omnia but we still wander if there is in fact demand for such devices. Is the market large enough and the area cool enough? Are there enough people who would value open hardware running open software even with a higher price tag? Any feedback would be most valued.
No (Score:2)
And by no, I mean yes. Is there space, sure, maybe consumer grade. Is it useful (beyond consumer grade? Probably not.
Re: (Score:3)
Re: (Score:3)
Well, to be fair, good networking would require an ASIC designed for networking. It may in fact actually exist, but I haven't found any (haven't looked) and there would be other parts that need to be open for me to consider it "Open Hardware". There are plenty of commodity parts to build a semi-open networking stack, but those are all closed source hardware components.
The question is, how "open" is open?
Re: (Score:2)
There's Openswitch: http://www.openswitch.net/ [openswitch.net] but I don't think the hardware is open. The example at LinuxCon was an Accton switch, controlled by OpenSwitch running on an embedded board.
Re: (Score:2)
IMO it's going to be hard to do any open network hardware going forward unless you start designing ASICs. Even gigabit routers (which are technically gigabit layer 3 switches) need an ASIC in order to fully saturate your WAN bandwidth if you are with a gig provider.
Re: (Score:2)
Even gigabit routers (which are technically gigabit layer 3 switches) need an ASIC in order to fully saturate your WAN bandwidth if you are with a gig provider.
Gigabit is reasonably trivial to do in software if you have a half-decent CPU. Get a OneAccess 1645 or a Juniper SRX220 and you are good to go. Yes, you can overwhelm at least the latter with 64-byte packets, but why would you do that? Both are using fairly lousy CPU's.
If you go Intel, it is difficult to find a processor slow enough that it would not route 1Gbps, even if you look at CPU's intended for tablets. Perhaps one of the really low-end phone processors.
Re: (Score:2)
Re: (Score:2)
No, technically they're routers (or gateways, if you want to use the IETF term). Switches are multiport bridges. "L3 switch" is a marketing term, created when wirespeed routing in hardware became available, in order to market them as having throughput comparable to switches.
Re: (Score:2)
"which are technically gigabit layer 3 switches"
No, technically they're routers (or gateways, if you want to use the IETF term). Switches are multiport bridges. "L3 switch" is a marketing term, created when wirespeed routing in hardware became available, in order to market them as having throughput comparable to switches.
What he said...
Layer 3 switches are technically routers. The difference is that a router is purposefully built with the necessary hardware to perform in hardware routing. Most switches that can perform layer-3 functions tend to do so in software or with the general purpose CPU. There are exceptions, of course, such as Cisco Supervisor blades for the 6500 switches which can have a MSFC with route processors, the PFC for netflow, etc. and the DFC with ASICS. However, they still tend to have a lower perfo
Re: (Score:2)
No. A router interconnects connects broadcast domains, i.e. it works at layer 3. It doesn't matter if the routing occurs in hardware or software. An "L3 switch" is market speak for a router which can work at (close to) wirespeed, although in most cases is very limited in the protocols it can support (try to find a wirespeed DECnet router).
Multiport bridges are called switches. The Kalpana was
Re: (Score:2)
"Most switches that can perform layer-3 functions tend to do so in software or with the general purpose CPU"
Unless they happen to be based around something like the Broadcom Trident2 or newer.
You're talking seriously old silicon that can only operate at L2
$60 4x100Mbps Open Switch got crowdfunded (Score:2)
There was a recent crowdfunding campaign for a open-protocol switch (I forget if it was OpenNFV, OpenFlow, or OpenVSwitch? Probably NFV.)
4-port 100 Mbps, so easy enough to do cheaply these days. I didn't really have any experiments I wanted to do with one that I couldn't also do with a virtual switch, so I didn't join the crowdfunding, and for production work I'd want at least GigE, but it was still interesting thing to go by.
Re: (Score:3)
Re: (Score:2)
I'm sort of going to second this too, I have an OpenRD which is using a slightly older version of the Marvell CPU. The docs initially appear open, but I quickly ran into a number of cases where they wanted me to sign an NDA to get the full documentation covering the information I needed.
Basically, all the useful information is not in the public tech docs...
Re: (Score:1)
Their previous offering, "Router Turris," despite having a reference design and being claimed as "open source hardware", used the Qualcomm QCA8337N-AL3C, so it isn't really open hardware, either, and is much more closed than hardware which doesn't require binary blobs to work (well).
You dismiss the previous boards as not being open hardware solely based on your conviction that the QCA8337N-AL3C requires binary blobs to run. However, this is not true, just check with the OpenWrt people.
Re: (Score:2)
thats the real question.
how is your "open" thing more open than a pc with bunch of network cards?
this sounds just like a slashvertisement, probably because it is. "would you buy our product?"
Re: (Score:2)
"Broadcom has the lion's share of the market for network ASICs, and is very much a closed environment. "
In the last decade they've gone from being sued for breaching GPL to being one of the most prolific contributors to the Linux kernel and have put a bunch of other stuff into GPL or BSD licensing.
As long as Avago don't screw the pooch, it's entirely possible that they will move towards being more open on asics.
Yes, please! (Score:2)
Particularly with the FCC racing to lock down router firmware, the market needs a player who will do the minimum the law requires, but provide as much freedom to tinker as possible.
Re: (Score:2)
Re: (Score:2)
Of course, but for a manufacturer, this means "sign our updates and allow nothing else".
Re: (Score:2)
Re: (Score:2)
"You can even get gigabit speeds on old retires desktops..."
Yes, at 70-90W power consumption, not 5-15
Re: (Score:2)
"You can even get gigabit speeds on old retires desktops..."
Yes, at 70-90W power consumption, not 5-15
Or the "cheaply available hardware" also mentioned at 5-15 watts... Talk about cherry picking.
Re: (Score:1)
I'll likely get one for VPN. OpenWRT forums (Score:2)
I'd be likely to get one. Maybe to use as a router, but I'd think of it more as a general purpose small network appliance to be cast into whichever role I need. Right now, I use a consumer router with *wrt as my VPN endpoint for the rescue network on my sever rack. It provides access to IPMI and the IP-controlled power strip through a VPN. (Meaning it's not used often, and doesn't need to be fast). Your device looks good for that type of purpose. Rack ears, preferably 0-U rack mounting, would be handy
Re: (Score:2)
Particularly with the FCC racing to lock down router firmware,
Which is a damn good reason to separate the router and the WiFi. The FCC can not do shit about http://www.smallwall.org/ [smallwall.org] or http://www.pfsense.org/ [pfsense.org] or any other router that works better than most commercial offerings on old cheap retired desktops.
Re: (Score:2)
http://tech.slashdot.org/story... [slashdot.org]
http://tech.slashdot.org/story... [slashdot.org]
false premise (Score:5, Insightful)
Raspberry Pi is not an open, depends on closed source blobs in firmware and drivers. Stop spreading the lie
Re: (Score:1)
Indeed. The Rpi is about as open, hardware-wise, as a gl-inet router (http://wiki.openwrt.org/toh/gl-inet/gl-inet). The only thing that makes these systems "open" is that they run Linux, give you root access right from the start, and encourage people to make their own software applications, OS, and hardware peripherals. Granted, this is much more than what one can expect from most manufacturers, but it doesn't make the hardware open. Still, looking at the apparent success of these 2 systems, I'd say there i
Re: (Score:2)
Re: (Score:2)
Nobody points to a typical PC and says "THIS IS AN OPEN ARCHITECTURE". But for some reason they lie about the pi
Re: (Score:2, Informative)
Raspberry Pi is a GPU with a CPU bolted on the side. The CPU is entirely dependent on the GPU, it cannot even boot without it. The GPU driver is fairly trivial; it mostly consists of stub calls to the firmware blob.
The firmware blob is proprietary and closed, to the point that it even has restrictions on which instructions the GPU is allowed to use. You can buy an unlock code to remove the restrictions. Even Android devices are not that closed down in general.
Re: (Score:3)
If you are referring to the MPEG decoding stuff, there is no restriction in the firmware that prevents it from working. It is merely that the RPi's creators require you to buy a licence to unlock the user land software that can make use of it because that is the agreement they signed with the manufacturer.
Realistically, not even hard core guys like RMS care about CPU microcode and GPU firmware too much. Stallman's preferred Thinkpad X60 has an Intel CPU with non-free microcode, and an Intel GPU with non-fre
Re: (Score:2)
If you are referring to the MPEG decoding stuff, there is no restriction in the firmware that prevents it from working. It is merely that the RPi's creators require you to buy a licence to unlock the user land software that can make use of it because that is the agreement they signed with the manufacturer.
This is simply wrong. The user land software is the same, the firmware functions are unlocked cryptographically. If it had been user land, cracks would have been available within hours of the release of the software.
Who knows which firmware functions could be unlocked if you send the right cryptographic checksum to the firmware? Perhaps the send_ssh_key_to_NSA function? Unlikely, but it is also unlikely that anyone will ever find out.
Re: (Score:1)
Up isn't down. Red isn't blue. Black isn't white, either, but some shades of gray are closer to black than white. More choice, whatever the shade is gray, is always good. Stop spreading the counterproductive extremism. -PCP
Re: (Score:2)
Raspberry Pi is not an open, depends on closed source blobs in firmware and drivers. Stop spreading the lie
An Intel Atom motherboard is much more open, and MUCH more powerful. And more expensive. Cheap goes a long way, and that is a problem for the posters of this slashvertisement.
A lesson long ago taught and learned (Score:2)
Good engineering is all about cost effective solutions, not whiz bang technologies.
Translation:
Just because it's "cool" (read that as PC, opensource etc) doesn't make it good engineering. If it's bad engineering (not cost effective), there is no place for it. period. If your price is higher because you think it's cool, go work for apple.
Re: (Score:2)
Just because it's "cool"
"cool" isn't quite the objective here.
(read that as PC, opensource etc)
In this case read it as:
secure, auditable, trustworthy...
If it's bad engineering (not cost effective), there is no place for it. period.
How do you measure "cost effective" in the context of releasing hardware that may have backdoors, security problems masked by non-disclosure, un-disclosed "features" the nominal "owners" find either undersirable or outright hostile, or even the simple risk of being held to the support or lack of support whims and pricing of the vendor?
Re: (Score:2)
Learn to quote, fucking idiot.
No. I really should learn to preview.
The issue is a simple typo; missing a / in the closing tag on one of the quote tag pairs. I'm as irritated as you by the result.
I think /. should bloody well report when the formatting tags aren't balanced when I hit submit; and then i'd fix them. It already checks and rejects a bunch of other things (too many capitals, too much punctuation, too frequent....) Blocking unbalanced quote tags should be a no-brainer.
Sure... (Score:2)
As long as it comes with service level agreements and support contracts and other such enterprise-related things.
Or as long as it's the cheapest thing you can get at your local big box retailer.
I'm here all night. Try the veal. (Score:2)
>but we still wander
Maybe you need a road map?
--
BMO
Re: (Score:2)
Copyediting. It's a lost art.
Re: (Score:2)
I was speaking about the world at large, really. I know Slashdot's never had copyediting.
Re: (Score:2)
I'm excited, at least (Score:1)
I saw the presentation at the OpenWrt Summit, and I got excited. I can't buy 10,000 units, but I'll probably start installing these by default wherever I can, assuming I can my hands on some at a price that works. I know I will run one at my house. =)
User opinion (Score:1)
Maybe (Score:2)
I think there is a discrepancy between your intended audience, home users, and the skills necessary to take full advantage of this platform. I could use it, but then I could build one of my own as well. General consumers want something that serves a well understood purpose and which requires little interaction. I think if you tailored it to a specific purpose, say as a security device which filters Internet traffic which was also you main WiFi access point it would sell. Being simply an open platform me
Selling point is missing. (Score:1)
One of the larger selling points of open hardware and software is the cost benefit. Almost every example of open vs closed system has the open system costing less initially. Now, hiring a team of dweebs to go to fix your father in law's wifi would be nice, but I would not expect the extra cost to cover that. Does it?
Wrong focus. (Score:1)
Open hardware and software is great and could be very useful and successful in the router space, but you end off by teasing high prices though you were using Raspberry Pi and Arduino for comparisons. Those products are designed to be straightforward and cost not much more than necessary so as to spread adoption. If you plan on charging extra for the open aspect of what you plan to make (or if the cost is due to making hardware to compete with the high end router market; completely unclear which it is with
Possible interest... (Score:2)
Our company provides networking services and we generally use Cisco gear, but we've been dipping our toes into some lower-end markets that can't afford $1,500 Internet routers. In order to consider something likes this, the main thing we would be interested in is build quality. It seems that most SOHO routers are designed right at the edge of their thermal safety envelopes, which leads to crashes and failures. Even if we don't want to spend $1,500 on a router we would still want something that is robust eno
Re: (Score:3)
Intel chipsets, gigabit speeds, and fanless. And runs the successor to m0n0wall, so it will run for years.
Or if you need more ports, they have more options.
You have to do it right (Score:2)
1. Poor price/performance: I'm using a used atom board that I got for $65 with 4GB RAM included. It's hard to beat stuff like that in cost-efficiency.
2. Proprietary cases and accessories: For the love of god, just make your board compatible with mini-ITX mounting holes so I can throw it in a plain old PC case if need be.
3. Not enough support for niche accessories: can it fit a huge Compex wifi card? All that would require in most situations is having one of the
Routers may be only defense left for privacy (Score:2)
Re: (Score:3)
With many modern OS's adding spying and telemetry features and then disabling all the tried and tested methods to bypass them it may wind up that the router is the only way to retain our digital privacy. So yes, I think open source networking has great utility.
True, but how is this better then the wealth of FOSS router projects our there now? SmallWall, t1n1wall, pfSense, OPNsense, BSDrp, OpenWRT, DDwrt, Untangle, or any *nix with routing turned on?
Yeah, a cheap 10 gig switch. (Score:2)
How long are they going to keep charging such ridiculous prices for 10 gig networking?
10 gig copper has been out now longer than it took 1 gig copper to go from being "ooh, enterprise" expensive to being in every $499 laptop you could find. Yet they've managed to prop up 10 gig switch and NIC prices forever.
Are 10 gig parts that complicated that they're staying so expensive for so long?
Or are we waiting for the next big "ooh, enterprise" speed bump to come along?
Re: (Score:3)
Are 10 gig parts that complicated that they're staying so expensive for so long?
Yes. 10Gbase-T runs a complicated encoding which requires power-hungry processing. Note how you cannot buy a 10Gbase-T SFP+ module for love or money; the power/cooling budget of SFP+ does not permit it to exist. Yet you can trivially get optical SFP+ modules designed for 80km, where 10Gbase-T is struggling to reach 100m.
10Gbps+ ethernet for the home will happen, but I doubt it will be 10Gbase-T.
Re: (Score:2)
Yeah, but that's one module type that tends to be a little unusual from a real-world usage scenario. Usually people who want 10G-T buy a switching platform that provides those ports built into the chassis or blade. People that buy chassis/blades with all SFP+ ports tend to use them as a fiber landing point or for terminating devices with SFP+ cabling.
The only place I see SFP-copper being used are on maxed-out blades or chassis where the only ports left are SFP and these tend to be 1G platforms anyway.
The
Re: (Score:2)
Until very recently, the cheapest option to get 10Gbps was SFP+ and Direct Attach Cables. It is still by far the most energy efficient option, unless you use 10Gbase-T with energy saving that lowers the port speed to 100Mbps when not much bandwidth is used (and that has latency implications).
I have not yet touched a single 10Gbase-T-port but 10Gbps SFP+ is everywhere. A few get actual fiber SFP+ modules inserted, but the vast majority is DAC.
Re: (Score:2)
Is energy efficiency that important in most use cases? Best I can tell, copper vs. SFP chassis differ by about 60 watts.for a 24 port model.
SFP+ may be cheaper but the cabling sure isn't and it makes it more expensive yet if you're needing to blend in a handful of 1G devices to the same stack.
Re: (Score:1)
Re: (Score:2)
How long are they going to keep charging such ridiculous prices for 10 gig networking?
10 gig copper has been out now longer than it took 1 gig copper to go from being "ooh, enterprise" expensive to being in every $499 laptop you could find. Yet they've managed to prop up 10 gig switch and NIC prices forever.
Which 10gig? There were several different ones, that have finally narrowed down to 2. With one of them being fiber... And...
Fiber sucks for endpoint networking. People step on network cords, tug on them, twist them around. Fiber can not do this.
And what are you going to push over it? Unless you have fast SSD, you can not even fill a gig. And to get full 10g, you need some super fast memory and buss speed as well as an SSD array.
Re: (Score:2)
10G is great for iSCSI storage and even all-hard disk arrays can overwhelm 1G links pretty easily. Multipathing helps, but like NIC teaming, doesn't result in anything like linear increases in throughput for additional paths.
A fair number of the storage installs I've done lately have been flash cached or tiered arrays and with 10G links, I've seen sequential throughput hit 600 MByte/sec.
Even if a given box can't completely saturate a 10G link, the gap between 1G and 10G is pretty wide -- I'd take 2-3G effe
Re: (Score:2)
Re: (Score:2)
To seriously saturate a 1gig link, you have to get into 5 figures. At that point, the price of 10gig-e is not a big deal.
Not a big deal until the customer starts comparing quotes. The uptick on the SAN controller for 10 gig ports, the uptick on any servers for 10 gig ports, and then the huge uptick on the 10 gig Ethernet switching.
The SAN controller price increase is of course way beyond the price increase of a generic two-port Intel 10G card, as is the OEM price increase on the server ports. The switchi
Re: (Score:2)
There was no use for gigabit ethernet in a consumer environment, either, but economies of scale won because it stopped making sense to produce 100 meg silicon.
My question is why is 10 gig silicon so much more expensive that manufacturers haven't given up on 1 gig silicon in their designs? Is it so vastly more complex that the cost increase outweighs the economies of scale?
Re: (Score:2)
Can do that now (Score:2)
I can make my own router out of existing open hardware like the RPi, but it's not cost effective to do so when I can buy a commodity home router for under $100USD (or $200 for a nice one) and have it last several years. It's a nice idea on paper, but it's just not practical given what the average consumer can get off the shelf today without the hassle of trying to build and configure the damn thing themselves. The parts alone are going to cost more than OTS routers today, and then you have to figure in the
Tor Node? (Score:1)
6 Pack (Score:2)
Look man, I dont even understand why (Score:1)
There is closed hardware in networking. Any monkey can do that. They are glorified old PC's and all their value is in the software anyway. Hell there is no such thing as a "hardware networking appliance". Those only exist in the imagination of people that think learning to configure an IP stack is a somehow complex ordeal.
Definitely yes... (Score:2)
So yes, please, build some open source hardware that will run tomato and/or OpenWRT (absolutely not DDWRT). I would l
Re: (Score:1)
Connecting 25 clients on the same access point at the same SSID, isn't that the rule of thumb people use when they design the deployment of WiFi in a sea of cubicles? That is the case with overhead antennas that make WiFi waves easier to send. Do not try
Explain the Value (Score:2)
Re: (Score:2)
5x Gb LAN ports
Gb WAN port
SFP port
2x Mini PCIe slots
mSATA slot
USB 3
4GB flash
1GB RAM
1.6GHz x2 CPU
"starting at $100 USD"
Good luck finding a Netgear router with a fiber connection for $100
Re: (Score:2)
pfSense (Score:2)
Recently, I picked up some used rack mount routers that are x86 based. Got them on eBay, plus upgraded the RAM to 2GiB, replaced CPU with one twice the clock speed and more cache, plus added larger storage. Installed pfSense on these things. Each one with upgrades was still under $100 thanks to people unloading seemly "useless" equipment on eBay!
These things are acting as more than just routers, too. Since it is FreeBSD based, there are tons of packages available. I've got nginx running on them as forward f
OpenWRT? (Score:2)
There is definitely room for open source routers. Buffalo has been in that market for a few years and as a result sell some of the most stable of routers I've seen outside the enterprise market.
If you're going to build this, make sure all is accessible including the radios, each port individually and if you're going into the enterprise market, extensions to OpenWRT for centralized management would be awesome. I'd also like to see a router with more than just 4 ports.
Almost there (Score:1)
What we have now in the marketplace is lots of routers with VLAN capable switches, which can all be running OpenWRT. For me it is not really important if the hardware is Open Source or not. I get a stable platform, and I can add USB and serial devices to it. And I can run whatever software on it that I want.
There are some complaints about the hardware accelerated switching whatever does not work well, but since since I am only on 60/60 mbps this is not big issue. I have a bigger problem with the built-in ss
optional 802.11ac (Score:1)
I am curious about everything related to the optional 802.11ac interface:
- chipset
- driver
- license of driver
- firmware
- performance
Other than this: very much yes.
I'll happily pay USD 100 for an open router platform. And pay extra for the wifi if.
Someone make a pure AP image with support for 802.11r/k and a hardware option for PoE, and you may have another winner.
OpenWRT Router in Nordic Countries (Score:1)
Open router platform (Score:2)
Let me know (Score:1)
...when there's an open source car.
Re: (Score:1)
Re: (Score:2)