Virtual Hosting 500 sites with per-site health monitoring

The obvious trick here is to do the bare minimum possible to keep it light and lean. Without knowing a bit more about your requirements it's somewhat tough to say what the best approach is. Here are a few follow-on questions:

 

 

1) For each virtual host, is there a different definition of "up"? For example, is one a static brochure site and the other a php-enabled one? It sounds like this is the case, but it's worth confirming.

 

 

2) If you want site-specific monitors, I don't see a particularly easy way around passing a HOST: header for each site you care about - after all, that's the only way to hook the virtual host container for each one. That means you're looking at adding lots of monitors to your config.

 

 

3) What is your fallback action? For example, if siteA is up and SiteB is down, what would you like to do with SiteB's traffic in response to that vhost being down?

 

 

Another possible approach could be to setup a basic monitor that gives you a 'good enough' idea of the overall health of the web server instance in question, then use an iRule with HTTP::retry, HTTP::respond, or HTTP::redirect to handle error conditions that you care to act on. This will likely accomplish what you want from a service perspective.

 

 

Check out this link for an idea of what is possible: http://devcentral.f5.com/Default.aspx?tabid=53&view=topic&postid=5201

 

 

Hope this helps.

 

 

-Matt