Hello guys, I came to here since I don't know what more I can do to my server. I've research a lot, including the manual, but nothing quite help me because of my server configuration and to something that I can't seems to identify.
This was not my build, I'm from support. Our server is virtual and has 64GB of vRAM and 24 vCPU. We have 348 sites on it. Other details:
* Server Version: Apache/2.4.43 (cPanel) OpenSSL/1.1.1g mod_bwlimited/1.4
* Server MPM: event
Our MultiPHP Manager configuration (We have tryied 500, 100, 100; this one is the default values):

Our Apache2 configuration:

I've made a Include of this value:

The thing is: our machine is not using a lot of RAM, it's using a lot of CPU (specially in rush hour). Right now, for instance, our load averages is 17.19 / 16.17 / 17.34.
Floating with it, it's the php-fpm which seems to be using a lot for each pool:


I've tryied to gather how much each process is using my vRAM with a Python script, here are the results:


16/17 vCPU been use is already noticeble in the performance in the website. They used to be around 8/9 vCPU usage, only now I'm facing this problem.
Dosen't matter what I tweak, this problem persist. Some pools is using a entire vCPU, and it's ever changing, I can't find a way to identify them.
Can someone enlight me if this tweak seems alright and what I need it's just a upgrade of my WHM server? It's important to say that, in rush hours, the vCPU usage goes crazy sometimes, hitting 26 to 30 load average. For example, as I writting this, I've press F5 in the Process Manager and some pools is hitting 160 / 650 CPU%, but is very quick; if I refresh again, it's gone.
Thank you for your help! If you need more information, let me know!
This was not my build, I'm from support. Our server is virtual and has 64GB of vRAM and 24 vCPU. We have 348 sites on it. Other details:
* Server Version: Apache/2.4.43 (cPanel) OpenSSL/1.1.1g mod_bwlimited/1.4
* Server MPM: event
Our MultiPHP Manager configuration (We have tryied 500, 100, 100; this one is the default values):

Our Apache2 configuration:

I've made a Include of this value:

The thing is: our machine is not using a lot of RAM, it's using a lot of CPU (specially in rush hour). Right now, for instance, our load averages is 17.19 / 16.17 / 17.34.
Floating with it, it's the php-fpm which seems to be using a lot for each pool:


I've tryied to gather how much each process is using my vRAM with a Python script, here are the results:


16/17 vCPU been use is already noticeble in the performance in the website. They used to be around 8/9 vCPU usage, only now I'm facing this problem.
Dosen't matter what I tweak, this problem persist. Some pools is using a entire vCPU, and it's ever changing, I can't find a way to identify them.
Can someone enlight me if this tweak seems alright and what I need it's just a upgrade of my WHM server? It's important to say that, in rush hours, the vCPU usage goes crazy sometimes, hitting 26 to 30 load average. For example, as I writting this, I've press F5 in the Process Manager and some pools is hitting 160 / 650 CPU%, but is very quick; if I refresh again, it's gone.
Thank you for your help! If you need more information, let me know!
Attachments
-
119.8 KB Views: 4
-
84.3 KB Views: 3
-
93.1 KB Views: 4