Hosting choices and speed test

[Adam Howard]

So I have finally narrowed it down to the following hosting service providers and I need YOUR help.

I know some of you have been around the block more than once and even doing my research, you may know something I maybe missing. So anything you can share would be helpful.

• www.webline-services.com
• www.crocweb.com
• www.vultr.com
• www.hostmantis.com
• www.linode.com
• www.steadfast.net
• www.quadranet.com

IF you do not know anything about any of the above, you can still be helpful by telling me which one of these is faster for YOU?

I have community forum filled with members all over the world, so my speeds isn't so important as getting a general average from others.

1. One
2. Two
3. Three
4. Four
5. Five
6. Six
7. Seven

^ The size of the download files are different so do not worry so much on how fast the file finished, but rather the average speed from the file. Technically, you do not need to finish the download, just let it run for 10 seconds and get an idea from that

 

[WSWD]

You do realize that tests like this are largely useless, and are just going to be dependant on location, right?

In any event, I have been using Quadranet for years, and HIGHLY recommend them. It's a very solid network.


Here are some speed tests for those files:

Dallas, TX





Connecting to webline-services.com|162.217.147.170|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 51041719 (49M) [application/zip]
Saving to: âtestfile.zipâ

100%[======================================>] 51,041,719  5.56M/s   in 19s




Connecting to www.crocweb.com|192.99.15.160|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: â100mbtest.binâ

100%[======================================>] 104,857,600 20.2M/s   in 6.3s




Connecting to wa-us-ping.vultr.com|108.61.194.105|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: âvultr.com.100MB.binâ

100%[======================================>] 104,857,600 27.1M/s   in 4.2s




Connecting to www.hostmantis.com|199.16.184.90|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: âtestfile100.binâ

100%[======================================>] 104,857,600 30.0M/s   in 3.9s





Connecting to speedtest.dallas.linode.com|2600:3c00::4b|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: â100MB-dallas.binâ

100%[======================================>] 104,857,600 56.4M/s   in 1.8s




Connecting to testfile.chi.steadfast.net|2607:f128:1:3::d|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: âdata.binâ

100%[======================================>] 104,857,600 47.0M/s   in 2.1s





Connecting to repos.lax-noc.com|2607:fcd0:0:a::2|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: â100mb.binâ

100%[======================================>] 104,857,600 38.1M/s   in 2.6s





Los Angeles, CA


Connecting to webline-services.com|162.217.147.170|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 51041719 (49M) [application/zip]
Saving to: âtestfile.zipâ

100%[======================================>] 51,041,719  15.4M/s   in 3.2s





Connecting to www.crocweb.com|192.99.15.160|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: â100mbtest.binâ

100%[======================================>] 104,857,600 17.2M/s   in 7.1s





Connecting to wa-us-ping.vultr.com|108.61.194.105|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: âvultr.com.100MB.binâ

100%[======================================>] 104,857,600 43.0M/s   in 2.3s





Connecting to www.hostmantis.com|199.16.184.90|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: âtestfile100.binâ

100%[======================================>] 104,857,600 30.8M/s   in 4.6s





Connecting to speedtest.dallas.linode.com|50.116.25.154|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: â100MB-dallas.binâ

100%[======================================>] 104,857,600 18.2M/s   in 5.7s




Connecting to testfile.chi.steadfast.net|208.100.4.54|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: âindex.htmlâ

100%[======================================>] 104,857,600 21.6M/s   in 5.8s




Connecting to repos.lax-noc.com|198.55.111.5|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: â100mb.binâ

100%[======================================>] 104,857,600  110M/s   in 0.9s

 

[TPerry]

I've been perfectly happy with QuadraNet (thanks @WSWD for the feedback on them). Since it's unmanaged I don't expect them to hold my hand. If I needed that I would expect to pay for their support tier (and so would they). For their unmanaged support the tickets I've entered have been quickly and concisely answered - but the answers usually require some finer knowledge of networking, etc. (for the questions I put to them) to understand what they are saying.

 

[Adam Howard]

You do realize that tests like this are largely useless, and are just going to be dependant on location, right?

YES and that is what I want

I have already determined that the host I'm looking into are all satisfactory.

However, as you have pointed out, location can be a factor. There have been a few times when I picked a good host that was fast for me, but everyone else couldn't reach it for beans (was slow for everyone else, everywhere else).

So the whole point of this is to see how well other people (not myself) can reach these hosting providers.

My members are from all over the world. In addition I wish to do the best I can to service, the world, and so the best way I can do that is to get an understanding on how good you can reach something from your part of the world.

The "average best" is what I'll likely go with.

But these threads also serve as a gateway for more opportunity to discover and investigate. Likely someone may know something about X, Y, Z hosting provider that I could not have learned otherwise. So this ends up being a "honeypot" to where I may hear an interesting story (good or bad).

So there is a method to my madness.

 

[WSWD]

I guess I'm still slightly confused then....

It's generally simple. Midwest (Chicago and Texas) are likely to get you the best average speeds for the US. East coast (New York) will be quicker to Europe. West Coast (Los Angeles) will be quicker to Hawaii, Japan, the rest of the Pacific, etc.

Also, you really should be asking for traceroutes instead of speedtests, because there are just too many factors involved with the speedtests. A few people downloading the speedtests at the same time and you are going to get way different results.

Traceroutes are going to get you generally the number of hops from one location to another, regardless.

 

[Adam Howard]

I guess I'm still slightly confused then....

It's generally simple. Midwest (Chicago and Texas) are likely to get you the best average speeds for the US. East coast (New York) will be quicker to Europe. West Coast (Los Angeles) will be quicker to Hawaii, Japan, the rest of the Pacific, etc.

Also, you really should be asking for traceroutes instead of speedtests, because there are just too many factors involved with the speedtests. A few people downloading the speedtests at the same time and you are going to get way different results.

Traceroutes are going to get you generally the number of hops from one location to another, regardless.

I know how to do a traceroute, but not everyone else does (least of all my members who are mostly not tech aware).

But traceroutes don't always determine speed either (surprised?).

I can do a trace on X domain and it could hop around through 30 different routes (no lie), but be faster than the one that only hop through 5. The route will even suggest the 5 has better speeds. Yet I'll download a 10 mb file that 30 route in 1 1/2 seconds, while the 5 may take 8 seconds.

If you can figure that one out... lol .... I'd rather keep it simple and just have people do speed test.

But you're not wrong about there being a wide range of other factors. Most of which you'll likely only know until you're actually on the host. But I would rather know as much as I could before getting there.

 

[WSWD]

And herein lies the problem. A shorter traceroute is ALWAYS going to yield a faster speed...always...theoretically of course.

All those "other factors" I talk about will change that. Port speed on the server, how overloaded the server is, how many folks are using the test download, etc. will all affect the speed. That's why they're largely useless.

 

[Adam Howard]

And herein lies the problem. A shorter traceroute is ALWAYS going to yield a faster speed...always...theoretically of course.

All those "other factors" I talk about will change that. Port speed on the server, how overloaded the server is, how many folks are using the test download, etc. will all affect the speed. That's why they're largely useless.

Not always (as I stated in my post above yours). But in theory it should.

This however is the best method that I can determine overall for everyone, without first hosting with someone... Whole point it to try to not jump from host to host or waste money opening accounts, just to find out.

There is no "perfect solution", there is only this one.

 

[Adam Howard]

In order to get a large polling and many results from many different people from all over the world, I posted this thread on several different sites. After counting the collective total (250 votes), I have narrowed down my list to these final 3 choices.

These are the providers who tested faster

• www.crocweb.com
• www.hostmantis.com
• www.linode.com

So we're down to the final 3.

 

[MattW]

Linode based on those above. That is from my own personal experience with Linode. @eva2000 is currently hosting with them, along with @DRE and a few other people who I've built a VPS for.

 

[Adam Howard]

Linode based on those above. That is from my own personal experience with Linode. @eva2000 is currently hosting with them, along with @DRE and a few other people who I've built a VPS for.

I love having my own setup. But I don't always have the time to manage it (I really don't). I also can't keep hiring someone anytime something goes "bump" (sorry).

So that is why I'm also looking at the Resell accounts at the other two.

Yet if I did go with Linode, I've saved the current nginx configuration files though from my current box, so I figure so long as I install Ubuntu on Linode, I should be able to just re-upload those and be all set.

decisions... decisions... decisions...

 

[Adam Howard]

This is the provider we decided on

referal link to our new provider https://my.hostmantis.com/aff.php?aff=603

 

[Jeremy]

The referral link argument does not belong in this thread. If you have a problem with a post, report it, don't argue over it in public.

 

[Adam Howard]

Linode based on those above. That is from my own personal experience with Linode. @eva2000 is currently hosting with them, along with @DRE and a few other people who I've built a VPS for.

This is what we're going to end up doing in July (2 weeks from now).

I have come to the unavoidable conclusion that we cannot cut corners as we once did by using a large reseller account. SU it would seem has "grown up" and truly does now demand a VPS or Dedicated Server.

Our reseller account is "manageable", but notably slower.