This is most likely your issue then.
The errors above around prefork workers means that you are getting more requests than Apache can handle with the current configurations.
Prefork is not the best mpm module to use, you should be using the event mpm module and then configure PHP to use php-fpm via proxy from Apache.
The apache event mpm might only need a small amount of configuration of its config file but best to test first then change only if you need to.
PHP fpm config would also need to be edited so that you are running the right amount of processes to handle incoming requests. Default is 5 I think, but the standard way is to work out how much RAM each process takes then divide RAM available to get the top value. You can use static or dynamic process handling, up to you. RAM available also depends on system RAM and RAM given to the DB and everything else.
Opcache might also need to be configured if it was only enabled, most likely giving more RAM but default might be ok for now. It can be tweaked later easily.
DB would need to be configed too, mysqltuner is a good start but make sure to research each recommendation as they may not work for your specific circumstances. That applies to everything though.
Also look into your MUC and sessions config. With a single server using SSDs there might not be much improvement, but you can use the tests available in the admin settings. Redis is the better option these days and I prefer two instances, one for MUC with an LRU policy and sessions without any purging. Redis is also only single threaded.
I would probably need to see the different config files to know more.