Personally, I now only buy and recommend routers that are supported by OpenWRT even if the end user does not end up using it. Usually such support means that the device is fairly open and uses fairly standard hardware. Something new and esoteric is likely to be more buggy. OpenWRT is also a great OS. If your router supports it, please give it a try, for security's sake.
I needed a new router recently to replace an Airport Extreme. I looked at gear supporting OpenWRT, Tomato, or DD-WRT. But I didn't want WiFi, and it was a PITA wading through the hardware lists.
I then came across the Ubiquiti EdgeRouter Lite. I had previous experience with Vyatta (of which EdgeOS is a fork), and was sold. I've been very happy with it.
Over a few dozen router purchases over the years, I've just settled on buying Buffalo DDwrt stock devices. If it runs that, it can run openWRT, and I'm happy. I don't have to dig around the openwrt hardware support forum or pray that I get the right revision of the model that works right.
And shipping foss firmware as the stock option wins my business. And they usually have ample onboard flash storage relative to their competitors for fun programs to run on the device.
That means the WRT1900AC is not one of your choices yet, right?
Also, I'm personally very bummed by the fact it has Marvell WiFi. They tend to be really mediocre and buggy, both in Windows, Linux and embedded applications.
I have been sticking to TP-LINK products lately. They and Buffalo seem pretty open and reasonably priced. Just a personal preference though. Lots of people I know are happy with Asus, Linksys, etc.
Also, from what I understand Microtik routers are a great alternative to OpenWRT, though I have not used them myself.
The mikrotik hardware is good but I find the documentation is poor and several times I have found their implementation of a feature is half baked. For instance their openvpn is over tcp only.
If you can make sure they do what you need and figure out how to set them up to achieve that they can be very cost effective. They are popular with the wireless ips crowd but have been facing strong competition from ubiquity.
Given the choice i buy ubiquity products, but they don't have as many "features".
I gave up on Mikrotik when I got one of their Switches, which explicitly have support for VLANs, and could only discover that the .. vlan..support was so botched that it couldn't make default VLAN's on different ports.
I think you have to "unslave" the ports from the master so they pass their packets up to the CPU, assign the ports to a bridge, and then the assign the bridge to a vlan.
It took me about 50 hours to figure out but I did manage to get three SSIDs on one radio, each of which would connect the client to a different vlan.
TP-Link WDR3600 is still the best value (non-AC), generally available, consumer grade dual band AP router on the market. Its ~impossible to get built-in modems with dual band Wifi on a functioning software stack, a separate HG612 (DSL) works great.
Microtik can be great value if you're willing to do lots of vendor QA (for other values of vendor) in your lab before going in to production. Most of the problems are to do with community: the docs are an unloved wiki which only staff can modify; release management, QA and bug tracking are third rate; and the forum is a passive-aggressive Engrish ghetto with poor moderation. But config management, CLI and Winbox are still better than straight Linux, and boards start at $40 which can make it worth it.
The WDR 4300 is on sale on newegg for under $50 after rebate, probably worth getting considering it's about the same price as the 3600 but has a slightly higher max bandwidth.
I assume you're judging based on client hardware? That may be an unfair comparison; AFAIK (working on enterprise wireless APs), they're mostly just known for catering to Cisco wireless AP radio chips (and doing well with those). Other markets are secondary.
Unfortunately, this means they don't care much about good drivers (including open source!) for anything other than Cisco's IOS.
Well, the large majority of home/SOHO APs/routers are Linux and my experience with Marvell-based devices in this segment was not good. I buy Atheros-based WiFi devices when I can.
Our startup has tried both the Asus RT68-U and the Netgear N7000 (the two top of the line, AC1900 class, consumer routers on Small Network Builder[1]) for both office and factory operations with both stock and DD-WRT[2] firmwares.
Our initial preference was to lean towards the Asus given their track record on OpenWRT for the past few years. If your last experience with flashing a router's firmware was with the Linksys WRT-54G back in 2004 than you may have missed out on all the Asus development that happened in the past decade.
We found the Netgear N7000 performed better overall. We were reluctant to use it as it did not support Dual-WAN mode initially like the Asus on stock firmware. (I.e. primary Comcast internet, secondary Verizon MiFi usb adapter.)
We recently updated both routers to [Kong]'s firmware mod of DD-WRT[3] and now the Netgear N7000 blows everything out of the water.
At our factory we are concurrently connecting 50 previously unassociated WiFi devices every 30 seconds and the dual-core 1Ghz Netgear just keeps on trucking with no problems. Our Asus meanwhile will crash about once a day.
Going forward we will stick with the Netgear. It seems to have become the 'default' router on the DD-WRT community, with support for other routers being forks off the N7000 code. Given Linksys has long no longer been 'Linksys' been 'Cisco' been 'Belkin' been Marvell, I don't anticipate it superseding the thrown despite the classic throwback black and blue livery. The 1.2Ghz processor and quad-band antenna is nothing to snub at. Should be an interesting year for router enthusiasts.
[2]In broad strokes, DD-WRT is like a nice, easy to install, easy to configure version of OpenWRT. If you previously preferred using Tomato to OpenWRT this is for you. OpenWRT is when you really want ring-zero console access only with no amenities out of the gate.
[3]http://tips.desipro.de/ - [Kong] made some modifications to DD-WRT to work better with the new AC-1900 class routers. They are fast becoming the master branch.
What build number of Kong's are you using for the R7000? I just picked one up to potentially replace a flaky RT-N16 running Kong but it's not the cheapest device for home use.
For that kind of money, it seems that Ubiquiti could be a viable option. Trying to decide if I should return the R7000 and get the Ubiquitis instead.
What's the model of router that people are recommending for Ubiquity? I recently upgraded my parents to an Asus and was extremely satisfied with Merlin's builds of the Asus firmware. Ideally, I just want something that runs Debian and lets me install packages so it can act as a home server. Why isn't this a thing yet?
I don't own one, but I usually see lots of raccomandations for the EdgeRouter Lite or the EdgeRouter POE (which is a Lite + integrated Switch + POE).
It does run on Debian, I think you can just ssh in an do everything you want. By the way, on DD-WRT/OpenWRT/Tomato you can install pretty much what you want by using OptWare/EntWare.
The problem is usually not the software, but the hardware. Router are still slow (for example, I tried once to use transmission on my puny e3200, and the speed was capped at ~350kbps by the CPU) and they don't have lots of connectivity (on high-end routers you get maybe 2 USB 3.0 ports. The WRT1900AC is in fact the first router with an eSata port, and the first one with decent speed over USB).
If you want to run anything somewhat "heavy", you're far better with having a different box, or using a custom pfsense/smoothwall/m0n0wall/vyatta box.
> The problem is usually not the software, but the hardware.
It's both. The hardware is slow and the software (in the cases you mentioned, the Linux network stack) is slow and inefficient.
The ERL uses hardware acceleration to avoid all this. It's capable of routing 1M PPS and the latest beta build does this with PPPoE as well. (In other words, it'll support Google Fiber connections well, for the lucky few.)
One thing to add is that the Edge Router Lite is based off of Vyatta and it is mostly feature compatible with it. It also boots off of a 4GB flash drive formatted as 2GB so if you open it up you can get double storage for free.
The firmware can be built with what we have provided, and will run on the WRT1900AC successfully. It is true that it would lack wireless support, but that would not stop the firmware from actually working.
> The code quality issues in the patches are fixable. The biggest problem with this is the fact that right now, the wifi chip (from Marvell) needs a proprietary driver to run. The submitted patches only include a prebuilt .ko for this driver. The response I got from Belkin indicates that they didn't realize that this was going to be a problem and they are now trying to fix it.
[That said, I find this to be typical of Belkin overall and would not, at the moment, recommend any of their Belkin/Linksys products if you expect to interface with them using software.]
I couldn't agree more. The Ubiquiti wifi hardware is better than any of the SOHO (Linksys, Asus, etc) I've ever used. I've changed most of my friends and family to using Ubiquiti stuff and I never hear about wifi problems anymore. The Nanostation and Rocket product lines are what I usually get.
I put 4 Unifi's in a friend's large house and it the performance is great.
I wish it had more than 4MB flash, but it is an awesome little box! The 4MB of flash just means you pretty much need to roll your own OpenWRT image if you want to change anything. A friend has one, and I have an MR3040 (3020 + 2000mAh battery) arriving today to hack around on to make it the perfect travel router for me.
Knowing this, can anyone enlighten me on why you'd buy this router over the Netgear R7000? It seems to be a worse router from the reviews I've seen, albeit with better storage performance.
Last time I tried a Netgear (a model from last year I think, AC enabled), it was awful. The web interface hadn't been updated since 2008-ish (could see stuff that was in older routers, old images etc) and didn't work properly.
Netgear support denied the existence of a telnet interface [1] and I returned it after some other problems I don't remember.
It still appears in their list of supported hardware though. http://wiki.openwrt.org/toh/start of course when you dig deeper it is clear it doesn't work but still should it be in one of the other tables showing support in progress
A great deal of performance differentiation in hardware comes from the drivers themselves, not the underlying hardware. This is evidenced in things like differing performance in similar situations using different OSes, such as a game in Windows vs. the same game in Linux (assuming they both us OpenGL).
The techniques that companies use to improve performance on their hardware are almost as complex and important as the hardware itself. This isn't to say that there isn't a benefit in open-sourcing drivers, just that companies are justified in being weary of open-sourcing their drivers.
The differentiation is more in microcode running on the radio chip than in anything that runs on the CPU. The microcode is what chip vendors are really paranoid about; I work on wireless AP firmware, and while we have the source code for all our wireless chips, vendors are very firm about never letting us see the source for microcode. If we need to debug a microcode issue, we file a case with them, and in emergencies they'll send an engineer on-site. We get binary blobs, the driver loads binary blobs, the driver sends config options for the advertised checklist of features to these binary blobs, and that's all we know.
The protectiveness of the driver seems more lawyer-driven. They're willing to hand out the source to pretty much everyone who asks to build a device, just under proprietary licenses.
Even when stuff does get open-sourced, the code takes quite a bit of work to pass muster for e.g. the Linux kernel. It's just so low-quality, regardless of vendor, that a lot of time and effort on the device-maker's end goes into debugging driver issues - work that is of course duplicated in a dozen or more companies. I've never personally looked at Marvell drivers, but OpenWRT's claim that it doesn't meet their standards is completely unsurprising to me.
To continue your point, often companies tier their products by their firmware to save development/manufacturing costs. If anyone can change the firmware, they may be able to make a $99 router perform like a $599 router (if I recall this was one of the big pulls for open-wrt or dd-wrt on the linksys 54g).
There is an upper limit of a piece of hardware, but if the software driving it's use is not exposing that or performing actions in a less efficient manner it can certainly have a negative impact.
Belkin effed up their firmware so badly on the F7D4301 that I got a new one for $35.
The original firmware was slow, buggy and required reboots every day. With Tomato it's one of the best deals I ever got :-) - dual band Wifi, Gigabit LAN, support for USB drives and 3G modems, no stupid blinking lights (I love that) - a perfect home/office router.
I a also guessing that most of the company that sell the hardware do not make it (just repackage it in a different plastic shell or wrap the drivers in an installer with pretty pictures). The hardware making company is likely not that much interested in giving away what it license to ASUS/Apple/Cisco etc....
> is evidenced in things like differing performance in similar situations using different OSes, such as a game in Windows vs. the same game in Linux (assuming they both us OpenGL).
I think better example would be the evolution of performance over different driver versions.
