Drupal - Guaranteed Failure
As you know, Drupal can't scale, can't handle any site with any significant traffic. Of course, some people will tell you that you can just use Varnish etc but that of course does not handle authenticated users. We will discuss how to fail (and how not to) with two strategies I have used to scale really big sites.
Everyone by now knows MongoDB (if you don't, you can fix that at http://munich2012.drupal.org/node/344 ). I will show how it integrates with Drupal, what scalability challenges it solves and what compromises you need to make if you use it.
I will give a quick introduction to Redis and show how to use it together with nginx and webdis to create a very fast Drupal site. Once again I will show what sort of sites it's a good fit for and what compromises you need to make.
Comments
The session name is http://team.onesite.com/crull/blog/2012/05/01/drupal---guaranteed-failure from here.
Hello chx,
good idea with this session title.
Just one link about "Managing site performance" for people landing here by search engine: http://drupal.org/node/627252
Last week there was an interesting session about Redis in our local php user group. Very nice but new technology for me. And now I see: There is more than one module for that. Your Redis SSI and some other modules. I love our community and most the very fast developers.
See you in munich.
Carsten
What is Redis SSI?
Can you list some more topics/details you plan to cover in this session?
The title sounds interesting and the one paragraph wets appetite. But I must say providing some more information would really help selecting this session.
@Stefan: The title and description of this content are a bit "tongue-in-cheek", i.e. it's clearly somewhat a joke. There was a recent head of a hosting company which apparently offers (and claims expertise in) Drupal services who had a bit of a melt-down and wrote a blog post saying that using Drupal would guarantee failure. My assessment was that using them to host a Drupal project was sure to help ensure failure, but of course there are many high-demand projects running on Drupal, so all they were doing was demonstrating how poorly equipped they are to properly manage a Drupal site. I'm sure we can learn a lot from someone as experienced as chx is about how to best negotiate the issues that lesser mortals might encounter. ;-)