[dangrossman]

You don't even need nginx or a cache. Bog-standard WordPress, no extra plugins, behind only Apache, will handle HN front-page traffic without falling over. All you need to tune is httpd.conf and my.cnf; the reason these sites die is that they allow more Apache processes to spin up than their little VPS actually has RAM.

The httpd processes push the MySQL posts table out of the OS's filesystem cache, the newly spun up processes eat up all the free memory, and you start swapping and the whole thing falls over. Turn off keepalive, set the maximum number of httpd processes to a number that leaves enough RAM for the DB and filesystem cache, and the blog would be fine.

My blog's been top of HN FP a few times. It's just apache/mysql/wordpress on the smallest Linode, sharing it with a half dozen other sites.

[fuzzix]

On something about half the spec of the lowest end Linode I'm not as worried about mitigating the HN effect (not something I have to worry about) as I am just getting the most I can from the instance - this means as few dynamically generated and database hitting requests as possible.

But yeah, like yourself, just pointing out that a Wordpress instance that can take a battering is eminently achievable.