Hi,
Your hardware is enough for 40-50k visitors a day. That is only a few requests per second on average. The problem is most likely configuration or uncached dynamic requests, not server size. I would not scale up yet. Find the bottleneck first.
1. Find out why it crashes
Crashes usually mean the server ran out of RAM. Check the system logs for out-of-memory kills, and watch CPU, RAM, and database activity during a spike to see what runs out first.
2. Use PHP-FPM with the event MPM
If Apache runs PHP in prefork mode, each worker can use 50-100MB. With a high MaxClients, a spike can use all 16GB. In WHM, switch to PHP-FPM in MultiPHP Manager. Then set the PHP-FPM max children based on your real memory use: the RAM you give to PHP divided by the average PHP process size.
3. Tune MariaDB
Set the InnoDB buffer pool to about 50-60% of RAM. Enable the slow query log and add indexes for the slowest queries. Slow queries are a very common cause of "slows down, then crashes".
4. Cache dynamic pages
Cloudflare does not cache HTML by default, only static files. If most hits go to PHP, it is not helping much. Add a full-page cache plugin and a Cloudflare rule to cache HTML for logged-out users. Add Redis for object caching if the site is database-heavy.
5. Check for bots
Look at your access logs for top IPs and user agents. Crawlers and attacks on login or XML-RPC endpoints can look like traffic spikes. Use Cloudflare rate limiting or WAF rules.
6. Nginx or LiteSpeed?
It can help, but only after the steps above. Apache with PHP-FPM and good caching handles this load fine.
7. Load balancing
This is overkill at your traffic level.
If you share your PHP-FPM settings, Apache worker limits, and slow query log results, people can give more specific advice.