Jump to content
Sign in to follow this  
MoonLiteNite

Fiber connection?

Recommended Posts

Hello!

I am currently away from my home, while on business. So i can't really test this, but dying to know...

 

I have had a 300/300mbs connection. And using speedtest.net it was hard to find servers that would go past the 100mbs upload speed, and only around 25% of the servers even hit the 300mbs download.

And about a week ago my ISP flipped the switch to turn on the 1000/1000/mbs connection!

 

I was wondering if anyone has had any luck hitting huge speeds on this??

I can't wait to get home in a week try it out!

Share this post


Link to post
Share on other sites

Ask sietec, he regularly pulls up to 900+ Mbps here. He's our speed all star around here but there are many more examples.

There are many factors that affect results here. Just keep trying to best your own speed... don't always expect to get 100% of ISP claims. Routing across the net needs work and depending on the servers and networks you connect to your results will be varied. TestMy.net tests are harder to ace too.

Having said that... TMN offers testing in many different ways. Single thread, multithread and to popular CDN services like Google PageSpeed and Cloudflare networks. To get a complete picture you need to try them all. Each test type provides different insight. Usually multithread is a good way to max out your line but that only tells part of the story. For some, multithread is slower. Really depends on the connection type and client machine.

I have something new I'd like to share with you.... but you have to wait a little longer. It puts the control of testing TOTALLY in the consumers hands. You can test from virtually any source, file type and file size on the Internet. The test looks and behaves just as you've come to expect from TestMy.net. Looks exactly the same except you can test against any almost any web server on the net. It's really easy too, input a url... that's it. I've seen up to 500 Mbps results with it so far. I'm going to open it up to allow even larger test sizes to help people with 1000+ Mbps connections pull longer tests... which allows the connection to fully ramp up and will yield higher accuracy.

I've tested hundreds of servers. From Google, Apple, Amazon, Facebook... nearly every server I've tried it on works perfectly and the results are verified by network interface readings. This update is already live and actually has been built into TMN's code for at least two years..... maybe up to 4-5 years, I really don't remember. I just forgot about it until I recently found a reference to it in the code. "Wait a minute..." --- searching for the variable I noticed turned up tons of results across my entire structure, deeply engrained into the program. It instantly worked and was highly accurate so I've build a database driven input system around it. I imagine that we can build a list of hundreds of thousands of test servers in short time. You can find the servers that work best for you and compare the results to my preferred test servers/methods.

I really want to just share it with you now. But I need to put the finishing touches on it and I'm driving through Pennsylvania heading to CT and NY. I should be able to sit back at my console and finish up on Sunday the 7th... releasing on Monday Sept 8th. Open beta available to all.

Giving you full control over where you test will give you a more realistic representation of your real speed. Using servers like Google, Amazon.. etc you can once and for all show your provider, "These are my speeds to very standard services. If you can't deliver my promised speed to these ultra powerful servers I'll find someone who will! Look, other people are able to pull faster speeds, why can't I..."

I think you guys will dig it. I hope that it helps people understand their bandwidth better and puts pressure on the ISPs to deliver what they claim... across the board not just to limited locales like is often the case today.

I aim to start revolution. But I always need your help spreading the word. Tell your friends! This site 100% relies on word-of-mouth.

I'll update this topic when it's released. Coincidently happens to be right about the time you get home. ;)

I already used this to prove to Softlayer that one of my servers had degraded bandwidth. Within minutes the issue was resolved. I told them, "Look, I'm testing against Google and Amazon servers that I was getting 400+ Mbps to earlier in the day... they are now testing at 60 Mbps. I've tested across at least 20 different networks and the results are the same." --- they never admitted that there was ever an issue but mysteriously it was resolved... all results returned to normal within about 20 minutes. During the time speed was degraded for speed tests to that server too, as soon as the results returned to normal using the new method my results to that server doing normal tests returned to normal as well. Again, they deny that there was ever an issue. TestMy.net PROVED otherwise. First day after I got a domain for 'it' I needed 'it' and 'it' proved itself. To me, 'it' clearly showed a bottleneck affecting all routes.

Soon I'll let you know what 'it' is and you can try 'it' yourself.

Share this post


Link to post
Share on other sites

Well, I have sad news, or at least up front it is sad.

According to testmy.net My upload is only at 22mbs or so.

Have run the test a few times now, and stays around the same.

 

According to speedtest.net... (the first server i selected... which i KNOW isnt a good one)

http://www.speedtest.net/my-result/3736832980

 

And according to torrents, I can easily upload a bit over 1,000Mb/s was hitting 1,100 again and again.

 

After a quick search of this site, i was unable to find the method to test many servers at once, in attempt to bog down my connection, where is that test located at?

Share this post


Link to post
Share on other sites

Your bandwidth is not as good as you think it is. You may be able to pull faster speeds but how long does it take to get there? It just like how most cars will do 120 mph but it's really all about acceleration.

Like I said, I have an upcoming option that will allow you to test from you own sources. If you don't trust my servers you can test off amazon, google or whatever you want. Trust me... using that you'll quickly see that bandwidth is not delivered equally to all the sites you visit.

You're getting killer numbers, don't get hung up that it's not as high as your ISP quotes. You're scoring higher than 99% of people here. The upload test right now only uses a max of 33MB of data and it's a single thread. Your score is great for that kind of test. I'm working to bring larger sizes and multithread to the upload test btw. Larger tests will bring higher number for insane connections like yours. Go upload a video to youtube and then tell me that you're still seeing 1000 Mbps. I bet money it falls much much much closer to the results you see here.

The tests here are not intended to help you achieve maximum speed. They're intended to tax your connection.

When you do a multithread test here you are combining the power of all the servers you select. Without selecting Google and Cloudflare that's 7000 Mbps of available bandwidth, much more when combined with the CDN options. I have more than enough bandwidth to serve you. If you were limited by my resources you'd get nearly the same speed on all your tests. Your fluctuating quite a bit.

Share this post


Link to post
Share on other sites

Your bandwidth is not as good as you think it is. You may be able to pull faster speeds but how long does it take to get there? It just like how most cars will do 120 mph but it's really all about acceleration. 

- Not really I load up an old torrent file that i have for a 40gb file, and within 30secs my upload speed is a bit over 1,000mbs. Same thing for downloads while on a privet tracker, only takes a few moments to shoot up.

- Sending files from my home PC, located in Austin, TX, to Santa Clara cali, over VPN, hits around 300mbs upload speed.

 

 

The upload test right now only uses a max of 33MB of data and it's a single thread. 

- Yup, i saw this, I figured this was the reason the upload sucks, by the time the upload can even start, it is finished, doesn't seem like a large enough chunk of data to get a good reading on faster connections.

 

Without selecting Google and Cloudflare that's 7000 Mbps of available bandwidth, much more when combined with the CDN options

- Just tried to disable those, and no luck. Mixed around with servers and not much change :(

 

I do understand how not every connection my computer makes to the world will be 1000mbs speeds. But i CAN see that when i am pushing data to dozens of users at once, i can each those numbers, so the bandwidth is indeed being provided by my ISP.

I would just like to see a 1 on 1 connection and have it actually use the full bandwidth, that is at least the feeling i get with speedtest.net, which i know, uses servers that are MADE to give you the best connection so it makes their service look good. 

 

edit:

Funny enough, doing the speedtest.net tests to the AT&T servers, my download is under 100mbs, but to some data enters up in kansas, or washington state, it shows 700 right now. I would think AT&T would inflate my numbers, but i guess not.

Share this post


Link to post
Share on other sites

...you have to connect to dozens of peers to see your full speed. That's not realistic Internet transfer outside of p2p. I bet you'll never upload to a single source at your full speed. Even awesome commercial 1000 Mbps will have a hard time doing that out to the Internet.

Here's an analogy. A hundred pennies is still a dollar but I'd rather have a dollar bill in my pocket. ... uploading or downloading from dozens of peers and combining the speed isn't the same as being able to do that from a single source. Some providers actually shape the way you use your bandwidth so that it's only possible to get full speed when you transfer many streams. Especially true with super fast residential connections.

Like I said, the upload test here wasn't designed with connections as fast as yours in mind. Because the test size is so small the best I've seen is a little over 300 Mbps. The download test is fully capable of gigabit speed however. Testing over LAN I've seen just shy of 1000 Mbps, about 990 Mbps (accounting for overhead, right on the money) ... it IS possible to get better speed here than you're getting. Look at sietec's results .. proof right there https://testmy.net/quickstats/sietec

I'd really expect that your multithread results would be higher. Be interesting to see what you get with the new test. I'll let you know as soon as I release it.

post-2-0-46301000-1410032392_thumb.jpg

Share this post


Link to post
Share on other sites

Yup, i fully understand that the odds of me actually ever downloading to my full bandwidth on a normal website is slim to none, at least until there is a true reason for servers to actually do it!

 

Like i was saying though, i just was hoping to find a speed test to actually show the bandwidth :)

 

Thanks for all your input, i will use this site from time to time to check speeds, and ill keep a lookout for your system to directly test speeds from the large websites.

 

Thanks!

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Sign in to follow this  

×
×
  • Create New...