The data genius behind Facebook's rise

Beijing Time February 5 News, Facebook has developed into the world's largest social network, with more than 1 billion active users. Over the course of nine years of growth, Facebook has not only encountered common problems in the development of the general company, but also because of the huge number of users, encountered more unimaginable "big data". What are some of the top data geniuses and state-of-the-art data technology support on Facebook's rise? The following is the full text of the article:

Jay Paric (Jay Parikh) 's desk is located in the 16th building of Facebook headquarters in Menlo, Calif., and his administrative assistant Gigny Samuel (Genie Samuel) works near him. Every once in a while, Patrick hears Samuel giggling, which means she's probably uploading some embarrassing photos on Facebook and tagging Patrick out. Usually, Patrick's notice on his Facebook page always follows her laughter. And if the notice does not arrive, he may be busy.

Patrick is the vice-president of Facebook's Infrastructure Engineering department, responsible for the hardware and software of the world's largest social network. If the notice on the page doesn't appear directly within a few seconds, then he will have to double-check his work and try to find out why. The trouble is that Facebook's infrastructure has been extended to four different places in the world, four large data centers, hundreds of thousands of computer servers, and, beyond that, more software tools-take a deep breath and it's hard to say their names without interruption. The reason for the failure of the notice is buried in a corner of the network.

That's why Patrick and his team will build tools like scuba. Scuba is a landmark software platform that can help Facebook engineers to dig and analyze the company's infrastructure data in a timely fashion. In general, it takes time to process such massive amounts of data. You may even need a few hours to finish everything. However, this is no longer a problem for scuba, a (in memory), memory-type data store that stores all of the data in hundreds of computer servers in high speed memory (rather than hard drives)-meaning you can query data more in real time.

"It gives us an idea of how the infrastructure works-how our servers are doing, how our networks are performing, how different software systems interact," says Patrick. So if Gini marks me in a photograph and the notice doesn't appear in a few seconds, we can look at the scuba to find the problem. ”

Nine years ago, Mark Zuckerberg, who launched the world's most popular social network in the world, Zuckerberg Facebook in his Harvard dorm. Zuckerberg and his company built the most advanced engineering business on the planet, and interestingly, the main reason they did so was because of the necessity. Facebook faces a unique and daunting task-serving 1 billion different users and dealing with 1 billion different messages, photos, videos, and many other data sources-a task that is more technically demanding than you might think.

Yes, there are many of Facebook's engineering armies, like Lars Rasmussen (Lars Rasmussen), who created the genius of social search tool graph Search and the people who helped beautify tools, test, and deploy applications. Now, Facebook's hardware engineers also have geniuses like Amir Michael (Amir Michael) who are responsible for server design, storage devices, and the entire data center management.

Facebook has also created a data-processing team of top engineers-the importance of this work in modern networks is becoming increasingly apparent. Scuba is just the tip of a number of Facebook's big data software platforms--using distributed systems that allow hundreds of thousands of computers to work together to disperse a single task.

Engineers, including Lagou Murti (Raghu Murthy), Avery Qing (, Avery Ching), Josh Metzler (Josh Metzler), built these excellent tools. Not only do they eliminate the problems in Facebook's data centers, but they also help faccebook data scientists, analyze the effectiveness of Facebook's online apps, and how users behave.

Although the "Big data" platform of Google (Weibo) is still recognized as the most advanced in the whole network. But as Facebook began to expand its Internet empire, the social-networking giant was not far away from Google, and, unlike Google, Facebook wanted to share its software with the world. Google is always sharing its ideas and ideas, while Facebook is sharing its own code and hoping others will take advantage of it. "Our company's mission is to connect the world and become more open," Patrick said. And while we are building our infrastructure, we are helping to promote that mission. ”

(Responsible editor: The good of the Legacy)