Centmin Mod vs Fastpanel

would need to know exact vps serve hardware specs, plan/web host and geolocation of both centmin mod and fastpanel servers as well as what types of page load tests you doing.

also test with webpagetest.org as well WebPerf - PageSpeed - How to use webpagetest.org for page load speed testing

if you're on a vps cloud provider, also do repeated tests with fresh new vps launched again, sometimes the vps host node you're on might be congested and more overloaded than another vps host node the other vps is on.

also double check the cpu and hardware of both vps are the same, some providers have different cpu/hardware for each vps host node so can make difference too

 

??? you have info we don't ?

 

don't think that is the case

@olivergreen also which version of centmin mod 123.08stable or 123.09beta01 ?

 

Same. though there's many LEMP/LAMP and mixed variant offerings out there these days ! Though not many that are totally free (and without paid support levels offered) with the level of support I provide via official community forums here

In general I don't mind folks pointing out if one alternative web stack is personally tested to be faster than Centmin Mod as there are many factors that can go into those end results. By all means if one tests faster for your specific usage cases, then use that yourself. But be aware of how and what you test matters too. Otherwise, you'll be masking performance deficiencies that you haven't tested for that exist when you're using the web stack and probably don't know it.

But ideally I would like specifics on test environment, test configs, how it was tested. Like how I provided all that info with my benchmarks linked below.

• Nginx non-https tests Nginx - Nginx Static HTML Benchmarks - Centmin Mod vs Easyengine vs Webinoly vs VestaCP vs OneInStack
• Nginx https tests Nginx - Nginx HTTP/2 HTTPS Benchmarks - Centmin Mod vs Easyengine vs Webinoly vs VestaCP vs OneInStack
• PHP-FPM tests PHP - PHP 7.x Benchmarks Centmin Mod vs Easyengine vs Webinoly vs VestaCP vs OneInStack

Clear example of how you test matters is look at results of Nginx non-https vs https tests linked above - different LEMP stack's Nginx server handled performance and scalability of non-https verus https tests differently.

• So if you only tested non-https only in your comparisons, you're missing out on https tests.
• If you only tested with HTTP/1.1 HTTPS testing tools, then you're missing out on HTTP/2 HTTPS testing results.
• If you only test static files (non-PHP), then you're missing out on PHP tests.
• If you only test cached PHP (i.e. fastcgi_cache PHP-FPM), then you're missing out on non-cached PHP performance - which is ultimately the real performance of your stack as for those 'cache misses'. Example if two LEMP stacks were tested, one had fastcgi_cache caching PHP-FPM requests and other didn't, of course fastcgi_cache PHP-FPM enabled LEMP stack would perform better for cached PHP requests. But what if you tested non-cached PHP-FPM performance and fastcgi_cache enabled LEMP stack did only 30 requests/s and the other non-fastcgi_cache LEMP stack did non-cached 100 requests/s. On cache miss under higher request load, which LEMP stack is better ?
• If you had one stack tested with varnish cache + nginx versus another with no varnish cache and just nginx, then you missed out on testing varnish cache'd 'cache misses' performance for non-cached requests. It's common to see folks say i installed varnish cache in front of web server, why under load it's still slow - due to 'cache misses' that hit origin backend web server and that web server not performing optimally.

I am sure with that testing/environment info, if Centmin Mod is indeed tested to be slower, with your/end user feedback I will be able to improve Centmin Mod to make it just as fast if not faster. That's the whole point of me developing Centmin Mod - to be the best LEMP stack that performs and scales optimally out of the box and have the optional features allowing it to scale and perform even further with custom tuning.

So by all means personally test all the LEMP/LAMP stacks out there yourself and see how Centmin Mod stacks up - yes pun intended I will never discount/ignore Centmin Mod users reports of slower performance compared to other LEMP/LAMP stacks if it's something I can improve/fix on Centmin Mod's default out of the box configurations. Centmin Mod isn't perfect, but with end user feedback/input/suggestions and reports, it can be made better

This of course excludes slowness due to needing to tune specifically for your web app/traffic - that would be something end users would need to deal with.

 

cheers @digij

Yeah pretty sure 90% of the time, if you claimed the same thing with other LEMP/LAMP stacks they'd either ignore you or get pretty defensive or somewhere in between. But for me, my mind is always open to the possibility that Centmin Mod isn't the fastest out of the box for all usage cases and that other LEMP stacks may be more optimally tuned for other usage cases.

So @olivergreen, let me know if that is the case and provided more config/host and test /environment info, I will try to work towards improving Centmin Mod even further