How We Made Github Fast (2009)

This blog post blew my mind. I generally thought of GitHub as being in the front-end business, but maybe they're really in the back-end business?<p>Is this sort of complexity common? I'm not saying I think it's bad, as I am sure it's all there for a good reason, but it just seems like <i>so many</i> services to bounce through before a web page is even served.

Ok! Now, make the website fast! I use GitHub for my documentation(wiki) and issue/bug tracker, and it takes ages to load those pages. I wanted to complain about it.. wrote a long feedback message, and it took 2mins to post it and then I saw the error GitHub page. How ironic.. The feedback system is too slow to post my feedback about the site being slow.<p>But, to be fair, it's not always like that. I guess that they had lots of trouble on our server recently and we've just been badlucky.

Nitpick: It's "GitHub", not "Github". Also note the date: <i>20 Oct 2009</i>.

What does he mean by "bare metal"? I've heard of attempts by Oracle to run directly on hardware, without an OS; does the same exist with mySQL and nginx? Or is he simply meaning not-a-VM?

This is an older post, but it's a high quality one to read.

Page did not respond in a timely fashion. Check our status site for alerts.<p>Either you found a page that took too long to render or we're getting more requests right now than we can handle.<p>You can try refreshing the page, the problem may be temporary. Learn how to deal with GitHub outages and other access problems.

What surprised me was that they were using 15k rpm drives over SSD's for database lookups. Any idea why?

Can someone explain how you do mysql replication via DRBD?

(2009)

"...The Aristocrats!"