up down up down

My textdrive websites (including this blog ) have been going down as much as they have been staying up. For no reason, whatsoever. As you can imagine, this is immensely frustrating. At some point in the past I setup Daedalus to detect the absence of a lighttpd process and restart it, but i realized today that the daedalus process wasn’t setup as a cron job, so it wasn’t running, thus my lighttpd instances weren’t getting restarted. I’ve been restarting them manually lately, whenever i discover the sites are down, which seem to pretty much daily. Why lighttpd suddenly ceases to run, I have no idea. I have lighttpd setup as a cron job to start every time the server boots, but this apparently insures nothing. Add to that the fact that the server – Pendrell – seems to cycle all the time, and goes down a lot. It was down yesterday. Now that Daedalus is cron’d, hopefully things we’ll be up a bit more!

We’ll I got a second account at Dreamhost, because A – they are cheaper, B – they support rails, C – why not, they had a sale. I have followed their instructions and got rails up and running. One thing that puzzles me though, despite the fact that my sites are running in production mode on Dreamhost, it seems like it is reloading everything on every request. (Unlike lighttpd on textdrive, i dont have to start any process, you just drop the right files in the directory, give the public folder execute permissions, and you’re off to the races.) BUT, is it caching anything? Keeping any ruby processes alive? If I change a file, i dont have to stop/start, it simply sees the change. And seems a bit slower than textdrive, like its reloading every time. But this contradicts the idea of production mode, where things are supposed to be cached. Is there a way to get things caching in memory, and boost the speed a bit? I hope someone can enlighten me.

Comments are closed.