Druid实时大数据分析原理与实践 - (EPUB全文下载)

文件大小：9.51 mb。
文件格式：epub 格式。
书籍内容：

版权信息
书名：Druid实时大数据分析原理与实践

作者：欧阳辰等

出版社：电子工业出版社

ISBN：978-7-121-30623-5

定价：79.00
版权所有·侵权必究
Foreword
Like many popular open source projects，Druid was initially created to solve a problem.We were trying to build an interactive analytics UI at a small advertising technology startup in San Francisco，and struggled to find a technology that could rapidly aggregate，slice and dice，and drill down into massive data sets.Eric Tschetter started the first lines of Druid to tackle this challenge，and that work has somehow led to an international community forming around the project.
I joined Eric on Druid soon after the project started，and for a while，the Druid world consisted of only 2 engineers.The first version of Druid was extremely minimalistic；there was a single process type，the“compute”node，and a handful of queries，but the core that was there was just enough to solve the problems with scale and performance we had at that time.
Our Druid cluster in the early years was less than 20nodes，and we worked around the clock to aggressively develop features and fix bugs.There were a lot of late nights in those days.I can still very clearly recall waking up in the middle of the night to fix an outage，and occasionally cursing loudly because the only reason the pager went off was because it was out of batteries.
As Druid matured，and as data volumes grew，we continued to face challenges around performance at scale and operational stability.Running in the then notoriously finicky Amazon Web Services cloud environment wasn’t always easy，and led us to make the decision to break up“compute”nodes into different components so that individual components could be fine tuned at scale，and any one component could fail without impacting the functionality of the other components.I am glad we made those decisions because it led us to sleep much more at night.
It has been extremely rewarding to watch the grassroots growth of the open source community.Unlike other popular open source projects，Druid was not developed at a major technology company or famous research lab.We open sourced the project without much attention，and the first open source version of the project almost didn’t have querying capabilities.We weren’t allowed to open source many pieces of the codebase，including most of the queries we developed.The night before officially announced the project，Eric was up writing GroupBy queries in a hotel room just so people could have a way of getting data out of Druid.After we released Dr ............

书籍插图：

以上为书籍内容预览，如需阅读全文内容请下载EPUB源文件，祝您阅读愉快。