Bigtable

BigTable是一種壓縮的、高效能的、高可延伸性的，基於Google檔案系統（Google File System，GFS）的數據儲存系統，用於儲存大規模結構化數據，適用於雲端計算。

BigTable發展於2004年^[1]，現今已成為Google的應用程式。像是MapReduce就常透過BigTable來儲存或更改資料，^[2]其他還有Google Reader^[3]、Google Maps^[4]、Google Book Search、"My Search History"、Google Earth、Blogger.com、Google Code hosting、Orkut^[4]、YouTube^[5]以及Gmail^[6]等。Google自行發展出特別的巨型資料庫的原因，自然是效能的問題^[7]。

BigTable不是傳統的關係型資料庫，不支援JOIN這樣的SQL語法，BigTable更像今日的NoSQL的Table-oriented，優勢在於擴充性和效能。BigTable的Table資料結構包括row key、col key和timestamp，其中row key用於儲存倒轉的URL，例如www.google.com必須改成com.google.www。BigTable使用大量的Table，在Table之下還有Tablet。每一個Tablets大概有100-200MB，每台機器有100個左右的Tablets。所謂的Table是屬於immutable的SSTables，也就是儲存方式不可修改。另外Table還必須進行壓縮，其壓縮又分成table的壓縮或系統的壓縮。客戶端有一指向META0的Tablets的指標，META0 tablets保儲所有的META1的tablets的資料記錄。

註釋

^ "First an overview. BigTable has been in development since early 2004 and has been in active use for about eight months (about February 2005)." Google's BigTable （頁面存檔備份，存於互聯網檔案館）
^ "Bigtable can be used with MapReduce, a framework for running large-scale parallel computations developed at Google. We have written a set of wrappers that allow a Bigtable to be used both as an input source and as an output target for MapReduce jobs". pg 3 of "Bigtable: A Distributed Storage System for Structured Data", 2006
^ "Reader is using Google's BigTable in order to create a haven for what is likely to be a massive trove of items." Official Google Reader （頁面存檔備份，存於互聯網檔案館） blog.
^ ^4.0 ^4.1 "There are currently around 100 cells for services such as Print, Search History, Maps, and Orkut." Google's BigTable （頁面存檔備份，存於互聯網檔案館）
^ "Their new solution for thumbnails is to use Google’s BigTable, which provides high performance for a large number of rows, fault tolerance, caching, etc. This is a nice (and rare?) example of actual synergy in an acquisition." YouTube Scalability Talk （頁面存檔備份，存於互聯網檔案館）
^ "How Entities and Indexes are Stored - Google App Engine - Google Code". [2011-04-05]. （原始內容存檔於2011-10-06）.
^ "We have described Bigtable, a distributed system for storing structured data at Google....Our users like the performance and high availability provided by the Bigtable implementation, and that they can scale the capacity of their clusters by simply adding more machines to the system as their resource demands change over time...Finally, we have found that there are significant advantages to building our own storage solution at Google. We have gotten a substantial amount of flexibility from designing our own data model for Bigtable." from the Conclusion of "Bigtable: A Distributed Storage System for Structured Data", 2006

外部連結

Bigtable: A Distributed Storage System for Structured Data（頁面存檔備份，存於互聯網檔案館） -（official paper; PDF（頁面存檔備份，存於互聯網檔案館））
BigTable: A Distributed Structured Storage System（video（頁面存檔備份，存於互聯網檔案館））
- more video
- Google's BigTable -（notes on the official presentation）
"How Google Works"^{[失效連結]}
Is the Relational Database Doomed ?

[1st-blog-1] "First an overview. BigTable has been in development since early 2004 and has been in active use for about eight months (about February 2005)." Google's BigTable （頁面存檔備份，存於互聯網檔案館）

[2] "Bigtable can be used with MapReduce, a framework for running large-scale parallel computations developed at Google. We have written a set of wrappers that allow a Bigtable to be used both as an input source and as an output target for MapReduce jobs". pg 3 of "Bigtable: A Distributed Storage System for Structured Data", 2006

[google-reader-3] "Reader is using Google's BigTable in order to create a haven for what is likely to be a massive trove of items." Official Google Reader （頁面存檔備份，存於互聯網檔案館） blog.

[maps-orkut-4] 4.0 ^4.1 "There are currently around 100 cells for services such as Print, Search History, Maps, and Orkut." Google's BigTable （頁面存檔備份，存於互聯網檔案館）

[5] "Their new solution for thumbnails is to use Google’s BigTable, which provides high performance for a large number of rows, fault tolerance, caching, etc. This is a nice (and rare?) example of actual synergy in an acquisition." YouTube Scalability Talk （頁面存檔備份，存於互聯網檔案館）

[6] "How Entities and Indexes are Stored - Google App Engine - Google Code". [2011-04-05]. （原始內容存檔於2011-10-06）.

[7] "We have described Bigtable, a distributed system for storing structured data at Google....Our users like the performance and high availability provided by the Bigtable implementation, and that they can scale the capacity of their clusters by simply adding more machines to the system as their resource demands change over time...Finally, we have found that there are significant advantages to building our own storage solution at Google. We have gotten a substantial amount of flexibility from designing our own data model for Bigtable." from the Conclusion of "Bigtable: A Distributed Storage System for Structured Data", 2006

[1]

[2]

[3]

[4]

[5]

[6]

[7]

Bigtable

相關條目

註釋

外部連結