By krozinov on

Hello folks,

I'm new to Drupal, and I've been looking over its capabilities and the forums over the last few days and have several questions that hopefully you can answer.

My biggest concern with Drupal is its scalability and performance.

Let me describe to you the type of site that we would like to potentially use Drupal for:
- Traffic of approximately 40 million page views per month (within weeks of launch). That's averages about 15 pageviews/second.
- Approximately 1 million users.
- The site will be Digg-like in nature: individual "story pages" each with user comments, voting on comments, ability to upload videos & photos.
- The site will have user profiles that will contain things like stories that the user has "dugg" and his/her friends.

This site's purpose is NOT to duplicate or clone Digg at all, just make use of some of it's cooler features.

Considering the functionality (modules) described above, I've got several questions:
1. How easy is it to install and configure Drupal to work in a webserver farm environment (multiple PHP/Apache/Drupal machines and multiple MySQL database machines)?
2. Can Drupal itself successfully handle (will pages load in about 1 second?) such a heavy load?
3. What kind of hardware specs do you think we would need to have Drupal successfully handle (will pages load in about 1 second?) such traffic?
4. Which version of Drupal would you recommend I use? I'm thinking the 5.x is the most stable and most feature-rich at this point?
5. Should I be concerned about anything else with Drupal?
6. Should Drupal be even considered?

I have looked at the Drigg module (Drupal+Digg), but I'm not sure how well it performs, but I'm guessing it all goes back to Drupal.

Thanks for any help or suggestions!
Konstantin

⋅ Categories: Drupal 5.x

Comments

iandickson’s picture

iandickson commented

There are plenty of huge sites using Drupal, hunt down the IBM tech docs released a while back.

I don't think that Drupal will be your problem IF you put it on the powerful enough hardware.

Ian Dickson

Likal.com

petervandijck’s picture

petervandijck commented

Are you launching a new site (and hoping for these numbers), or moving an existing site (and you *know* these are the numbers)? If it's the first, you shouldn't worry about scaling so much. Good luck!

petervandijck’s picture

petervandijck commented

And to answer the question: yes, of course Drupal can handle 40 million pageviews per month. It all just depends on how much hardware and stuff (memcached etc) you put behind it.

delorie’s picture

delorie commented

Looking back at your #5 question: "Should I be concerned about anything else with Drupal?"
Yes, especially about security issues, D/DOS attack, fraud, brute force etc, because you're running a famous website. Competitor will hold you down I guess =) .. hope not

kbahey’s picture

kbahey commented

40 million pv/mo means 1.33 million pv/day.

Drupal can handle that for sure. I know sites that can handle 1 million a day on a single server.

But it all depends on several things.

1. How much registered users are logged in at the same time on the site. The more, the more load on the database. Non logged in users are easy, and caching helps a lot, e.g. memcache or boost.

2. The exact modules you are using, and how "heavy" they are on the database. Some modules are detrimental to performance and can really bog down a site.

I recommend that you start as a minimalist: use as few modules as necessary. Do not get into a module binge, then complain about performance being bad.
--
Drupal performance tuning, development, customization and consulting: 2bits.com, Inc.
Personal blog: Baheyeldin.com.

--
Drupal performance tuning and optimization, hosting, development, and consulting: 2bits.com, Inc. and Twitter at: @2bits
Personal blog: Ba