1 / 34

It’s all about SCALE!!

It’s all about SCALE!!. How to scale up web service in the past ?. Source: http://www.slideshare.net/mallipeddi/scalable-lamp-development-for-growing-web-apps. 大型網站所使用的工具. Perlbal - http://www.danga.com/perlbal/ 多個網頁伺服器的負載平衡 MogileFS - http://www.danga.com/mogilefs/ 分散式檔案系統

teige
Download Presentation

It’s all about SCALE!!

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. It’s all about SCALE!!

  2. How to scale up web service in the past ? Source: http://www.slideshare.net/mallipeddi/scalable-lamp-development-for-growing-web-apps

  3. 大型網站所使用的工具 • Perlbal - http://www.danga.com/perlbal/ • 多個網頁伺服器的負載平衡 • MogileFS - http://www.danga.com/mogilefs/ • 分散式檔案系統 • 有公司認為 MogileFS 比起 Hadoop 適合拿來處理小檔案 • memcached - http://memcached.org/ • 共享記憶體?? • 把資料庫或其他需要經常讀取的部分，用記憶體快取(Cache)方式存放 • Moxi - http://code.google.com/p/moxi/ • Memcache 的 PROXY • More Resource: • http://code.google.com/p/memcached/wiki/HowToLearnMoreScalability • http://www.slideshare.net/techdude/scalable-web-architectures-common-patterns-and-approaches

  4. Source: http://mashraqi.com/2008/07/memcached-for-mysql-advanced-use-cases_09.html

  5. Source: http://www.slideshare.net/northscale/moxi-memcached-proxy

  6. Source: http://www.slideshare.net/northscale/moxi-memcached-proxy

  7. 教育訓練課程 HBase Intro 王耀聰 陳威宇 jazz@nchc.org.tw waue@nchc.org.tw

  8. HBase is a distributed column-oriented database built on top of HDFS.

  9. HBase is .. • A distributed data store that can scale horizontally to 1,000s of commodity servers and petabytes of indexed storage. • Designed to operate on top of the Hadoop distributed file system (HDFS) or Kosmos File System (KFS, aka Cloudstore) for scalability, fault tolerance, and high availability. • Integrated into the Hadoop map-reduce platform and paradigm.

  10. Benefits • Distributed storage • Table-like in data structure • multi-dimensional map • High scalability • High availability • High performance

  11. Who use HBase • Adobe – 內部使用 (Structure data) • Kalooga – 圖片搜尋引擎 http://www.kalooga.com/ • Meetup – 社群聚會網站 http://www.meetup.com/ • Streamy – 成功從 MySQL 移轉到 Hbase http://www.streamy.com/ • Trend Micro – 雲端掃毒架構 http://trendmicro.com/ • Yahoo! – 儲存文件 fingerprint 避免重複 http://www.yahoo.com/ • More - http://wiki.apache.org/hadoop/Hbase/PoweredBy

  12. Backdrop • Started toward by Chad Walters and Jim • 2006.11 • Google releases paper on BigTable • 2007.2 • Initial HBase prototype created as Hadoop contrib. • 2007.10 • First useable HBase • 2008.1 • Hadoop become Apache top-level project and HBase becomes subproject • 2008.10~ • HBase 0.18, 0.19 released

  13. HBase Is Not … • Tables have one primary index, the row key. • No join operators. • Scans and queries can select a subset of available columns, perhaps by using a wildcard. • There are three types of lookups: • Fast lookup using row key and optional timestamp. • Full table scan • Range scan from region start to end.

  14. HBase Is Not …(2) • Limited atomicity and transaction support. • HBase supports multiple batched mutations of single rows only. • Data is unstructured and untyped. • No accessed or manipulated via SQL. • Programmatic access via Java, REST, or Thrift APIs. • Scripting via JRuby.

  15. Why Bigtable? • Performance of RDBMS system is good for transaction processing but for very large scale analytic processing, the solutions are commercial, expensive, and specialized. • Very large scale analytic processing • Big queries – typically range or table scans. • Big databases (100s of TB)

  16. Why Bigtable? (2) • Map reduce on Bigtable with optionally Cascading on top to support some relational algebras may be a cost effective solution. • Sharding is not a solution to scale open source RDBMS platforms • Application specific • Labor intensive (re)partitionaing

  17. Why HBase ? • HBase is a Bigtable clone. • It is open source • It has a good community and promise for the future • It is developed on top of and has good integration for the Hadoop platform, if you are using Hadoop already. • It has a Cascading connector.

  18. HBase benefits than RDBMS • No real indexes • Automatic partitioning • Scale linearly and automatically with new nodes • Commodity hardware • Fault tolerance • Batch processing

  19. Data Model • Tables are sorted by Row • Table schema only define it’s column families . • Each family consists of any number of columns • Each column consists of any number of versions • Columns only exist when inserted, NULLs are free. • Columns within a family are sorted and stored together • Everything except table names are byte[] • (Row, Family: Column, Timestamp)  Value Column Family Row key TimeStamp value

  20. Members • Master • Responsible for monitoring region servers • Load balancing for regions • Redirect client to correct region servers • The current SPOF • regionserver slaves • Serving requests(Write/Read/Scan) of Client • Send HeartBeat to Master • Throughput and Region numbers are scalable by region servers

  21. Regions • 表格是由一或多個 region 所構成 • Region 是由其 startKey 與 endKey 所指定 • 每個 region 可能會存在於多個不同節點上，而且是由數個HDFS 檔案與區塊所構成，這類 region 是由 Hadoop 負責複製

  22. 實際個案討論 – 部落格 • 邏輯資料模型 • 一篇 Blog entry 由 title, date, author, type, text 欄位所組成。 • 一位User由 username, password等欄位所組成。 • 每一篇的 Blog entry可有許多Comments。 • 每一則comment由 title, author, 與 text 組成。 • ERD

  23. 部落格 – HBase Table Schema • Row key • type (以2個字元的縮寫代表)與 timestamp組合而成。 • 因此 rows 會先後依 type 及 timestamp 排序好。方便用 scan () 來存取 Table的資料。 • BLOGENTRY 與 COMMENT的”一對多”關係由comment_title, comment_author, comment_text 等column families 內的動態數量的column來表示 • 每個Column的名稱是由每則 comment的 timestamp來表示，因此每個column family的 column 會依時間自動排序好

  24. Architecture

  25. ZooKeeper • HBase depends on ZooKeeper (Chapter 13) and by default it manages a ZooKeeper instance as the authority on cluster state

  26. Operation The -ROOT- table holds the list of .META. table regions The .META. table holds the list of all user-space regions.

  27. Installation (1) 啟動Hadoop… $ wget http://ftp.twaren.net/Unix/Web/apache/hadoop/hbase/hbase-0.20.3/hbase-0.20.3.tar.gz$ sudo tar -zxvf hbase-*.tar.gz -C /opt/$ sudo ln -sf /opt/hbase-0.20.3 /opt/hbase$ sudo chown -R $USER:$USER /opt/hbase $ sudo mkdir /var/hadoop/ $ sudo chmod 777 /var/hadoop

  28. Setup (1) $ vim /opt/hbase/conf/hbase-env.sh export JAVA_HOME=/usr/lib/jvm/java-6-sunexport HADOOP_CONF_DIR=/opt/hadoop/confexport HBASE_HOME=/opt/hbaseexport HBASE_LOG_DIR=/var/hadoop/hbase-logsexport HBASE_PID_DIR=/var/hadoop/hbase-pidsexport HBASE_MANAGES_ZK=trueexport HBASE_CLASSPATH=$HBASE_CLASSPATH:/opt/hadoop/conf $ cd /opt/hbase/conf$ cp /opt/hadoop/conf/core-site.xml ./$ cp /opt/hadoop/conf/hdfs-site.xml ./$ cp /opt/hadoop/conf/mapred-site.xml ./

  29. Setup (2) <configuration> <property>    <name> name </name>   <value> value </value>  </property> </configuration>

  30. Startup & Stop • 全部啟動/關閉 $ bin/start-hbase.sh $ bin/stop-hbase.sh • 個別啟動/關閉 $ bin/hbase-daemon.sh start/stop zookeeper $ bin/hbase-daemon.sh start/stop master $ bin/hbase-daemon.sh start/stop regionserver $ bin/hbase-daemon.sh start/stop thrif $ bin/hbase-daemon.sh start/stop rest

  31. Testing (4) $ hbase shell > create 'test', 'data' 0 row(s) in 4.3066 seconds > list test 1 row(s) in 0.1485 seconds > put 'test', 'row1', 'data:1', 'value1' 0 row(s) in 0.0454 seconds > put 'test', 'row2', 'data:2', 'value2' 0 row(s) in 0.0035 seconds > put 'test', 'row3', 'data:3', 'value3' 0 row(s) in 0.0090 seconds > scan 'test' ROW COLUMN+CELL row1 column=data:1, timestamp=1240148026198, value=value1 row2 column=data:2, timestamp=1240148040035, value=value2 row3 column=data:3, timestamp=1240148047497, value=value3 3 row(s) in 0.0825 seconds > disable 'test' 09/04/19 06:40:13 INFO client.HBaseAdmin: Disabled test 0 row(s) in 6.0426 seconds > drop 'test' 09/04/19 06:40:17 INFO client.HBaseAdmin: Deleted test 0 row(s) in 0.0210 seconds > list 0 row(s) in 2.0645 seconds

  32. Connecting to HBase • Java client • get(byte [] row, byte [] column, long timestamp, int versions); • Non-Java clients • Thrift server hosting HBase client instance • Sample ruby, c++, & java (via thrift) clients • REST server hosts HBase client • TableInput/OutputFormat for MapReduce • HBase as MR source or sink • HBase Shell • JRuby IRB with “DSL” to add get, scan, and admin • ./bin/hbase shell YOUR_SCRIPT

  33. Thrift $ hbase-daemon.sh start thrift $ hbase-daemon.sh stop thrift • a software framework for scalable cross-language services development. • By facebook • seamlessly between C++, Java, Python, PHP, and Ruby.  • This will start the server instance, by default on port 9090 • The other similar project “rest”

  34. References • <趨勢科技>HBase 介紹 • http://www.wretch.cc/blog/trendnop09/21192672 • Hadoop: The Definitive Guide • Book, by Tom White • HBase Architecture 101 • http://www.larsgeorge.com/2009/10/hbase-architecture-101-storage.html

More Related