Benchmark Script + Nice UI!

Hi there,

I have just spent my last week and coded a nice benchmarking script, which tests your IO, network, gives information about the network, server hardware and more.

https://www.petabyet.com/

It was actually my first time to code an app completely with HTML/AJAX as the frontend, PHP/MySQL as the backend (and with help from people) so I'm pretty thrilled about the beta release.

Just go and click on the orange "run benchmark" box, you'll then be presented with a few questions, after that click "generate command", wait a few seconds for the AJAX/PHP and you'll get your command. The system actually prevents people from "cheating" by adding unique tokens to each download and also with other server-side authentication/verification algorithms. (make sure you turn javascript on when the system is generating your command). It'll take about 2-5 minutes to run, depending on your system.

Here is a sample report ran with my box @SkylarM (crissic) https://www.petabyet.com/result/55d44094cf1b729e7e1a6ca9cc6a773f/

Yes, the website is not quite finished yet, I'm here asking you to populate the database a bit and after that I can add more COOL things (yes, they are really cool) to the system, such as "PetaByet Score" which uses complex algorithms to make a 0-100 rating for your VPS, a database of VPS plans and various charts/metrics about the benchmarks.

Also, since the script/system is in beta stage, I'm here taking advices/suggestions as well as bug reports.

What do you guys think?

Thanks in advance

What do you have that ServerBear doesn't?

notFound said:

What do you have that ServerBear doesn't?

Click to expand...

Well, there is a custom CPU benchmarking script and a custom metrics script under development; there are also more speed test servers.

FYI: This script is not going to be a competitor of ServerBear or any other benchmark scripts, it's just another complement that you can choose.

Net said:

FYI: This script is not going to be a competitor of ServerBear or any other benchmark scripts, it's just another complement that you can choose.

Click to expand...

then I'll pass

I ran some tests on a VM at NodeDeploy, worked well i suppose.

Tried it twice and it timed out at the ThinkBroadband UK speed test both times.

Only thing I'd say is that rather than display the ISP on the homepage, you display the provider name.  For instance, we appear as ServerAxis.  This may lead to some confusion, especially when those 'ISPs' also provide VPS services. 

Thanks for the feedback!

Virtovo said:

Only thing I'd say is that rather than display the ISP on the homepage, you display the provider name.  For instance, we appear as ServerAxis.  This may lead to some confusion, especially when those 'ISPs' also provide VPS services. 

Click to expand...

The ISP display worked for me.  The Xen HVM VPS I tested doesn't have a "provider" .

Design is very nice. I never liked Serverbear, good to have a better alternative now!

Virtovo said:

Only thing I'd say is that rather than display the ISP on the homepage, you display the provider name.  For instance, we appear as ServerAxis.  This may lead to some confusion, especially when those 'ISPs' also provide VPS services. 

Click to expand...

Thank you, it has been fixed. I have also added the "PetaByet Score" section which can be found in individual reports. Currently the calculation is based on (Price/Benchmark/adjustments), suggestions are welcome .

Net said:

Thank you, it has been fixed. I have also added the "PetaByet Score" section which can be found in individual reports. Currently the calculation is based on (Price/Benchmark/adjustments), suggestions are welcome .

Click to expand...

That's much better.  I like how you kept the 'ISP' in there also.

Location: Fatal error: Uncaught exception 'InvalidArgumentException' w, Fatal error: Uncaught exception 'InvalidArgumentException' w, Fatal error: Uncaugh

Might wanna sort that out.

My VPSes are in the lead so far  :lol:

Good addition with the PetaByet overall score.

blergh said:

Location: Fatal error: Uncaught exception 'InvalidArgumentException' w, Fatal error: Uncaught exception 'InvalidArgumentException' w, Fatal error: Uncaugh

Might wanna sort that out.

Click to expand...

Could you please CURL the following URL's and post the result

https://www.petabyet.com/core/city.php

https://www.petabyet.com/core/state.php

https://www.petabyet.com/core/country.php

k0nsl said:

My VPSes are in the lead so far  :lol:

Good addition with the PetaByet overall score.

Click to expand...

Currently, the calculation (for network) is based on CacheFly only. Do you think it's a good idea to use the average of Internode CDN+Cachefly CDN or use an average of all locations?

@

Yes, I figured it went with CacheFly only. I think using an average score based on all locations would be best, but I am not so sure. Calculating only based on CacheFly would perhaps yield a more precise result.

I remain undecided on it, also very tired right now -_-

Btw, are you in #vpsboard or in #FranTech? Maybe we can talk some more about it tomorrow.

Currently, the calculation (for network) is based on CacheFly only. Do you think it's a good idea to use the average of Internode CDN+Cachefly CDN or use an average of all locations?

Click to expand...

Using an average would be better.  (my argument against a calculation based only on cachefly results is explained in )

DomainBop said:

Using an average would be better.  (my argument against a calculation based only on cachefly results is explained in )

Click to expand...

Changed it to the average of: Cachefly, Internode CDN, Softlayer San Jose, Softlayer Washington, i3d, Prometeus, Softlayer SG, ransom IT.

Now, if one of your network results is under 500kb/s, you'll never get a score of 5 or over.

k0nsl said:

Btw, are you in #vpsboard or in #FranTech? Maybe we can talk some more about it tomorrow.

Click to expand...

You can PM me

My only issue is the 'dd' benchmark; Benchmarking on VPS's with 'dd' is silly because the containers do not have direct access to hardware, the host node does some mmap() magic and other tricks plus some emulation of interrupts to<->from the containers.