Jump to content

Your servers cant handle my down/up?


patt2k
 Share

Recommended Posts

Seems strange that testmy servers cant handle my (75/35) in reality it's 84/35-39 mbps because of download fluff 

 

most I got was 61 mbps and even with big files its really random

while pretty much on every download file I hit 11.00 MB/s download 

 

any ideas why your DC servers cant handle those speeds?

Link to post
Share on other sites

Seems strange that testmy servers cant handle my (75/35) in reality it's 84/35-39 mbps because of download fluff 

 

most I got was 61 mbps and even with big files its really random

while pretty much on every download file I hit 11.00 MB/s download 

 

any ideas why your DC servers cant handle those speeds?

 

 

It's not that the servers can't handle the speed.  My servers are connected with 2000 Mbps of bandwidth.  Tier 1 bandwidth... Especially my main server in Texas... trust me, it's not that the server can't handle it.  It could handle 200 of your connections simultaneously if it needed to.

 

You may want to try testing with larger test sizes.  36MB is actually pretty small for your connection speed.  You may need more time to ramp the speed up.  Try 100MB.  ... could be a little slow down in the route too, but with Verizon I'd doubt it... Verizon normally peers very well.

Link to post
Share on other sites

First, you can't compare Ookla speed test results to TestMy.net speed test results.  Read Why Do My Results Differ From Speedtest.net / Ookla Speed Tests? to understand why that is.  Many providers use those tests because they make them look better...

 

Also, don't read too far into the location of the server.  As long as you've got clean routes and things are running as they should be you can max your connection across the country.  Testing with a server right next to you isn't really an accurate representation of your Internet speed.  When you surf the Internet your more often going to be hitting servers further away.  Your benchmark should be further away if you want to really put your connection to the test.  Providers like to test you as close to home as possible because it eliminates many variables that you're here to test... therefor your speeds will be higher.  That and Ookla tests also adjust the results to offset unknown variables... and muti-thread the test which can also lead to inflated results.  You shouldn't need to muti-thread to pull your speeds.  By their own admission Ookla drop the slowest 30% of the results... how can that be accurate?  The bottom 30% of the results is important information and is what you're here to see... right?

 

You can select Automatic Route from the top of the Speed Test pages to be routed through a CDN that will deliver the test from a location closer to you.  Choosing that option you should be routed through Newark NJ.  

 

I see now that you've found the option for Automatic Route... and your results are very similar.  How fast is your computer?  TMN can be as much of a test of your computers speed as it is a test of Internet speed once you start to reach the speeds that you're reaching.  Once you start getting into 50+ Mbps even your hard drive performance can be a factor because of the methods that TMN uses.  I've seen on more than one occasion where a computer on a 50 Mbps connection was only able to pull ~20Mbps... swap out for a faster HDD or SSD and BAM, full speed.  Other connection speed tests don't work this way.  TMN is more of a representation of your browsers performance coupled with your Internet performance.  If either is lacking in any way... your results will reflect that.

 

... but trust me... if things are as they should be, you can get some pretty insane results.  My servers are very fast.  The result below is testing at well over 2000 miles from the server...

 

HRzSivN.PgGLQcF.png

 

... you don't need to test on a server close to you to be accurate.  That's a misconception.

 A longer route, in my opinion, is a more accurate representation of your REAL speed.  If your provider can't deliver your speeds at a distance, what's the point?  The Internet is at a distance... it's the World Wide Web... not the New York Web.  Right?

Link to post
Share on other sites

Thanks for the reply again. I am still learning your site 

 

my specs are

 

14GB of ram DDR3 1600 MHZ

core i7 930

1X 80GB SSD for Windows

1X 250GB ssd

1X USB 3.0 , 2TB Drive

1X 3.0 GB/s 1TB Drive

Radeon HD 7950

 

so those specs are more then enough and yes I do know flash tests are not the best however I have another example , and trying to get better results here on your site :)

 

And please don't think your servers are crap from my point of view I just wanna know , this way maybe I will improve things on my side :)

 

let me know what I can do to improve results on your site :) because I acutally like the auto-tester to give me a lot of results. Thank you

 

second SS is from DD_WRT router

post-69832-0-77577500-1359168734_thumb.p

post-69832-0-89591000-1359169142_thumb.p

Link to post
Share on other sites

You've got an awesome computer... that's probably not the problem.

 

... try testing to the Central server.  That's the most powerful and best connected server I have.  It's right smack in one of the biggest Internet intersections on the planet.  Even though it's further away, you might get better results from there.

 

Is that screenshot a torrent?

 

Even if you can't achieve your full speed that doesn't mean that you can't still accurately benchmark here.  Compare your results to your previous results.  Using your Max speed that you've been able to reach as a benchmark to judge your performance on.

 

Realize that TestMy.net is less forgiving than other tests... and is a single thread test (unless your modem does channel bonding).  You might only be able to pull your full speed if you open more than one thread... which is what's happening when you download torrents or from usenet for example.

Link to post
Share on other sites

Ca3le, quick question. You posted earlier that ISP's want you test closer because it gives better results. In all actuality, they are only responsible for your bandwidth until you exit their network. If you have to hop through 15 connections to connect to another computer/server, your ISP is only going to be 3-4 hops. After that your at the mercy of whatever backbone you get routed to. So isn't it wiser to test closer to make sure the bandwidth your ISP provides is true? I mean, I could be getting 32 Mbits/sec on their network but hit a bad hop somewhere and barely get 5 Mbits/sec. That 27 Meg difference isn't my ISP's fault. Not to bash or troll, just want to clarify. 

Link to post
Share on other sites

Ca3le, quick question. You posted earlier that ISP's want you test closer because it gives better results. In all actuality, they are only responsible for your bandwidth until you exit their network. If you have to hop through 15 connections to connect to another computer/server, your ISP is only going to be 3-4 hops. After that your at the mercy of whatever backbone you get routed to. So isn't it wiser to test closer to make sure the bandwidth your ISP provides is true? I mean, I could be getting 32 Mbits/sec on their network but hit a bad hop somewhere and barely get 5 Mbits/sec. That 27 Meg difference isn't my ISP's fault. Not to bash or troll, just want to clarify.

Ah yes... But your provider does have control over that to a degree. They choose who they peer with. If they have crappy intermediate routes that they go through... Or the people they peer with do... Then actually, it is their fault. A provider with good peering that can route well across the net is a truly good provider. Not one that can only do that near by. No matter what they tell you... They DO have control over that. As much as you have control over your provider they do too, their peers are their providers. If one doesn't pull their weight, they shouldn't be using them.

Link to post
Share on other sites

Not to argue (well, discuss is more like it) but you said it yourself, they do have some control, to a degree. If you go across multiple backbones to get to a server, and there are issues between the 2nd and 3rd (like I have going from Tulsa to Cali), then there is little your ISP can help with that. They can control who they connect to, but not much after that. One single ISP can not control the routing for the whole network. In some cases every ISP in your local area take the same path at some point to a server. Even though they may have their own backbone, those backbones connect to a regional NSP then get routed the same. In my case here in Tulsa I know to avoid servers in Cali due to a backbone between me and them having load issues, one of the main reasons I always use testmy for speed tests at a distance. Even when I lived in Las Vegas, Cali was always a problem. Granted, I use the same ISP, but the trace's are much different (in Las Vegas I saw the issues with Big Iron, here in Tulsa its usually alter.net or xo.net that I see the lag jump).  Now if you have issues with an ISP across teh board, then yeah its time to switch ISPs. But if its just localized to one area, and especially if its after multiple backbones, then I would say its pretty much out of your ISP's hands. 

Link to post
Share on other sites

Not to argue (well, discuss is more like it) but you said it yourself, they do have some control, to a degree. If you go across multiple backbones to get to a server, and there are issues between the 2nd and 3rd (like I have going from Tulsa to Cali), then there is little your ISP can help with that. They can control who they connect to, but not much after that. One single ISP can not control the routing for the whole network. In some cases every ISP in your local area take the same path at some point to a server. Even though they may have their own backbone, those backbones connect to a regional NSP then get routed the same. In my case here in Tulsa I know to avoid servers in Cali due to a backbone between me and them having load issues, one of the main reasons I always use testmy for speed tests at a distance. Even when I lived in Las Vegas, Cali was always a problem. Granted, I use the same ISP, but the trace's are much different (in Las Vegas I saw the issues with Big Iron, here in Tulsa its usually alter.net or xo.net that I see the lag jump).  Now if you have issues with an ISP across teh board, then yeah its time to switch ISPs. But if its just localized to one area, and especially if its after multiple backbones, then I would say its pretty much out of your ISP's hands. 

 

i've experienced what your talking about aswell but generally it gets fixed with a week or month if its not getting fixed then someone isn't doing their job

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.

 Share

×
×
  • Create New...