Server load peaks randomly

troya

Verified User
Joined
Jul 8, 2019
Messages
9
Hello,

I have a server, migrated from cpanel and I never had any high load issues when I used to have cpanel. When I migrated first time, I tried openlitespeed. Server load has started to peak randomly. Then I changed to nginx+apache and it was quiet for some months. Few days ago, for some reason, the server load started to peak because of MySQL usage. It reached ~25. I couldn't find any reason when I checked.

Then, I changed to openlitespeed. and since then server load has started to randomly peak, same as the first time when I moved from cpanel. However, It does not stick as it did when nginx+apache. When I check with top, I don't see any process with high CPU utilization. Mostly I used to see lsphp processes. So, considering too many lsphp processes being opened, I reduced the child workers to 2 here:
/usr/local/lsws/conf/httpd-extprocessor.conf PHP_LSAPI_CHILDREN=2
/usr/local/lsws/conf/httpd_config.conf.in PHP_LSAPI_CHILDREN=2
/usr/local/lsws/conf/httpd-phplimits.conf PHP_LSAPI_CHILDREN=2

However, I still see the same random peaks. Nothing has changed.
I am sure I don't have even one website with big hits daily or big no of concurrent users. The websites on the server is more or less the same as when I used cpanel.

Now, I am thinking of this could be due to wplogin / xmlrpc attacks? I have followed some guides to reduce these attacks. Even reduced number of failed login attempts to #9 in Admin settings.

I don't want to go back to cpanel, as I made a big effort for everything to get working, and now I have got some experience with Directadmin.

I appreciate If you guide me for solving it.
 

Zhenyapan

Verified User
Joined
Feb 23, 2018
Messages
158
Location
UA
Check your website logs, I saw this problem on few servers simultaneously from last thursday - in my case it caused by increased googlebot activity, it's increased 5-8 times for some websites.
 
Top