High Performance at Massive Scale: Lessons Learned at Facebook | Hacker News Reader

High Performance at Massive Scale: Lessons Learned at Facebook

idleprocess.wordpress.com

66 pointsniels11 comments

Interesting quote from the article:

"Jeff also relayed an interesting philosophy from Mark Zuckerberg: 'Work fast and don’t be afraid to break things.' Overall, the idea to avoid working cautiously the entire year, delivering rock-solid code, but not much of it. A corollary: if you take the entire site down, it’s not the end of your career."

And apparently, don't worry about violating your user's data privacy!

I felt compelled to write this since I just received the Facebook Beacon class action lawsuit email a couple of minutes ago.

Why the downvotes?

More on Haystack: http://www.facebook.com/note.php?note_id=76191543919

The article mentions that Haystack has been open sourced but that doesn't seem to be the case yet.

A few days ago I sent a message to the FB developer that wrote the blog post about Haystack.

He told me that they are working on getting it open sourced and that it will be ready 'soon'.

wow. they have no (ie random) partitioning - it's just brute force. i hope their "Graph-based caching and storage systems" is coming along... :o)

[edit: although, given the exponential growth they give and the factor of 4 headroom they have left, they're going to top out in about 1.5 years anyway]

Yep, so basically one request can touch hundreds of servers. It's basically the opposite approach of Twitter. They aggregate the graphs on write.

Who is "They" in this case? Both a read and a write could be a "request", so I am having some trouble with the ambiguity of your last sentence. Thanks!

Twitter is they.

Which also is the reason you can only have 5000 friends on FB. If you had a million friends, you'd probably touch all their servers for one request.

An impressive statistic:

"In aggregate, Memcache at Facebook processes in 120M requests/sec."

Reply on news.ycombinator.com