Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Comcast internet speed improved with non-Comcast hardware

jeffreymartens.com

Post reply on HN
Print view View original

61–70 of 181 posts

Re: Comcast internet speed improved with non-Comcast hardware

#61

Post by noonespecial »

throwaway6556 wrote: post #23
manyxcxi wrote: post #15

The best thing anyone can do is swap out the junky cable modem + WiFi router combos they rent to you. They're horrible under standard usage, bad when lots of clients are connected, bad wifi range, low memory, require frequent reboots, etc. For about $70 (about 5mos of rental charges) you can get a Surfboard barebones Docsis 3 modem that will vastly outperform whatever junk they rent out. After that you could get a TP…

TP-Link is Chinese crap, please stop recommending it. Thanks. Ubiquiti is fine, and Debian-based, to boot.

Really? I've found the opposite. TP-Link has managed to match recent Ubiquiti's on performance and beaten them on price, but hasn't tried to ram a bunch of useless cloud crap down my throat when all I want is a simple AP.

Re: Comcast internet speed improved with non-Comcast hardware

#62

Post by newman314 »

manyxcxi wrote: post #15

The best thing anyone can do is swap out the junky cable modem + WiFi router combos they rent to you. They're horrible under standard usage, bad when lots of clients are connected, bad wifi range, low memory, require frequent reboots, etc. For about $70 (about 5mos of rental charges) you can get a Surfboard barebones Docsis 3 modem that will vastly outperform whatever junk they rent out. After that you could get a TP…

Protip: buy used off of craigslist.

I've done this multiple times over the years and saved a bundle.

Re: Comcast internet speed improved with non-Comcast hardware

#63

Post by johnbellone »

chrissnell wrote: post #34

Don't use Speedtest.net. The major providers all prioritize their traffic and it gives unrealistic results. I built a graphical console-based speed tester (client+server) that you can run yourself to get a more accurate picture and test between your home and servers you actually use: https://github.com/chrissnell/sparkyfish

I think that Google provides one if you search "Speedtest" on a non-mobile device.

Re: Comcast internet speed improved with non-Comcast hardware

#65

Post by johnbellone »

nodesocket wrote: post #22
johnbellone wrote: post #20

Earlier quoted context omitted.

I just decided to jump on the cord cutting bandwagon. My bill was ~$180 and I was able to get it down to $97 for the "best Internet". My cost for PlayStation Vue is $29.99 plus $15 for HBO. It streams to the PS4 and Amazon Fire TV. I am very happy with the change simply because I am not giving Comcast more money than I have to.

I need sports though, that's the problem.

PS Vue has sports-they are running channels specifically for basketball during march madness.

Re: Comcast internet speed improved with non-Comcast hardware

#66

Post by fake-name »

chrissnell wrote: post #34

Don't use Speedtest.net. The major providers all prioritize their traffic and it gives unrealistic results. I built a graphical console-based speed tester (client+server) that you can run yourself to get a more accurate picture and test between your home and servers you actually use: https://github.com/chrissnell/sparkyfish

Another reason to not use speedtest.net is they use compressible data for their tests.

DOCSIS has some support for stream compression, so you'll often get highly unrealistic results.

On the other hand, basically all internet content is sent over the wire already compressed, so any stream compression doesn't help with it at all.

I've had speedtest.net overrate my connection by 2-3 times it's actual throughput.

Re: Comcast internet speed improved with non-Comcast hardware

#67

Post by joecool1029 »

We had a Comcast SMC-D3G on Comcast Business. It's garbage. Once we pushed too many connections through it, it would crash and packet loss would skyrocket requiring a manual restart (unplug/replug). This is because it's also a very poor router that can't be changed to bridge-only. There's still plenty of these pieces of crap in the wild, and they also can't properly route ipv6.

If you're a business class customer that needs static ip's, you're stuck with Comcast hardware. Fortunately all my locations now can get away with dynamic ip's so we run our SB6121's (deprecated, but fine for 50/10 or 75/15 tiers).

The final bullshitty thing that's started in the past year is $100 per on-site visit if they can't prove the issue was caused by their network. I used careful wording there because if this is a peering/intermittent issue they'll come out next day just to pocket that charge and blame you.

Re: Comcast internet speed improved with non-Comcast hardware

#68

Post by gergles »

johnbellone wrote: post #28
joshfraser wrote: post #10

Another way to get the advertised speeds is to regularly call and complain. From my experience, they'll dial your speed up, just to stop you from calling in all the time.

You also need to keep in mind if you bought your own modem that you may need to upgrade the firmware (or worse case, a new modem) to get advertised speeds. This generally is advertised as 8x4 channels or better. My old surfboard tops out at 150Mbps.

You as the end user can't 'upgrade the firmware' in cable modems, it's all sent from the CMTS.

It's how they enforce "supported modems" - no firmware for your modem, you can't use it.

Re: Comcast internet speed improved with non-Comcast hardware

#69

Post by chrissnell »

fake-name wrote: post #66
chrissnell wrote: post #34

Don't use Speedtest.net. The major providers all prioritize their traffic and it gives unrealistic results. I built a graphical console-based speed tester (client+server) that you can run yourself to get a more accurate picture and test between your home and servers you actually use: https://github.com/chrissnell/sparkyfish

Another reason to not use speedtest.net is they use compressible data for their tests. DOCSIS has some support for stream compression, so you'll often get highly unrealistic results. On the other hand, basically all internet content is sent over the wire already compressed, so any stream compression doesn't help with it at all. I've had speedtest.net overrate my connection by 2-3 times it's actual throughput.

Interesting. I didn't know that. For what it's worth, sparkyfish uses randbo [1] to generate a fast random stream of data for the speed testing.

[1] https://github.com/dustin/randbo

Re: Comcast internet speed improved with non-Comcast hardware

#70

Post by wernercd »

throwaway6556 wrote: post #23
manyxcxi wrote: post #15

The best thing anyone can do is swap out the junky cable modem + WiFi router combos they rent to you. They're horrible under standard usage, bad when lots of clients are connected, bad wifi range, low memory, require frequent reboots, etc. For about $70 (about 5mos of rental charges) you can get a Surfboard barebones Docsis 3 modem that will vastly outperform whatever junk they rent out. After that you could get a TP…

TP-Link is Chinese crap, please stop recommending it. Thanks. Ubiquiti is fine, and Debian-based, to boot.

The TP-Link "crap" I've bought has been bottom priced and dependable.

Why by expensive routers, repeaters, etc when you can get a TP-Link for $20 that lasts for years and if it breaks... you spend another $20 to get the newest version... AND you've still spent less than the alternatives?

I do spend the extra where warranted... but TP-Link is top notch "crap" in my book.

Post reply on HN