I’ve seen this issue before, and it was due to not enough memory. Although it was a Windows Server, when MySQL reached its allocation of memory, the service shut down and needed a restarted.
More about your Hestia setup could help suggest some solutions. is it apache only, nginx, only or both? Are you using MariaDB or MYSQL? Are you using any caching?
Tweak your server, or look for a bigger spec server Looking in /var/log/syslog should show OOM errors (search for “OOM”) if I recal. I’ve had it before. Normally its from misconfigured max clients / memory per process / etc)