開源雲計算技術系列（六）hypertable（hadoop hdfs)

選擇virtualbox建立ubuntu server 904 的虛擬機器作為基礎環境。

hadoop@hadoop:~$ sudo apt-get install g++ cmake libboost-dev liblog4cpp5-dev git-core cronolog libgoogle-perftools-dev li bevent-dev zlib1g-dev libexpat1-dev libdb4.6++-dev libncurses-dev libreadline5-dev

hadoop@hadoop:~/build/hypertable$ sudo apt-get install ant autoconf automake libtool bison flex pkg-config php5 php5-cli r uby-dev libHTTP-access2-ruby libbit-vector-perl

hadoop@hadoop:~/build/hypertable$ sudo ln -f -s /bin/bash /bin/sh
[sudo] password for hadoop:

hadoop@hadoop:~$ tar xvzf hyperic-sigar-1.6.3.tar.gz

hadoop@hadoop:~$ sudo cp hyperic-sigar-1.6.3/sigar-bin/include/*.h /usr/local/include/
[sudo] password for hadoop:
hadoop@hadoop:~$ sudo cp hyperic-sigar-1.6.3/sigar-bin/lib/libsigar-x86-linux.so /usr/local/lib/

hadoop@hadoop:~$ sudo ldconfig

hadoop@hadoop:~/build/hypertable$

hadoop@hadoop:~$ wget HTTP://hypertable.org/pub/thrift.tgz
--2009-08-17 21:12:14--  HTTP://hypertable.org/pub/thrift.tgz
Resolving hypertable.org... 72.51.43.91
Connecting to hypertable.org|72.51.43.91|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 1144224 (1.1M) [application/x-gzip]
Saving to: `thrift.tgz'

100%[======================================>] 1,144,224   20.9K/s   in 44s

2009-08-17 21:13:00 (25.3 KB/s) - `thrift.tgz' saved [1144224/1144224]

hadoop@hadoop:~$ tar xvzf thrift.tgz

hadoop@hadoop:~$ cd thrift/
hadoop@hadoop:~/thrift$ ls
aclocal       config.guess  contrib       lib           NEWS              ylwrap
aclocal.m4    config.h      CONTRIBUTORS  LICENSE       NOTICE
bootstrap.sh  config.hin    depcomp        ltmain.sh    print_version.sh
CHANGES       config.sub     DISCLAIMER    Makefile.am  README
cleanup.sh    configure      doc           Makefile.in  test
compiler      configure.ac  install-sh    missing       tutorial
hadoop@hadoop:~/thrift$ ./bootstrap.sh
configure.ac: warning: missing AC_PROG_AWK wanted by: test/py/Makefile:80
configure.ac: warning: missing AC_PROG_RANLIB wanted by: test/py/Makefile:151
configure.ac:44: installing `./config.guess'
configure.ac:44: installing `./config.sub'
configure.ac:26: installing `./install-sh'
configure.ac:26: installing `./missing'
compiler/cpp/Makefile.am: installing `./ depcomp'
configure.ac: installing `./ylwrap'

hadoop@hadoop:~/thrift$ ./configure

hadoop@hadoop:~/thrift$ make -j 3

hadoop@hadoop:~/thrift$ sudo make install

hadoop@hadoop:~/thrift$ sudo ldconfig

hadoop@hadoop:~$ ls
hypertable-0.9.2.6-Alpha-src.tar.gz  jdk-6u15-linux-i586.bin
hadoop@hadoop:~$ chmod +x jdk-6u15-linux-i586.bin
hadoop@hadoop:~$ pwd
/home/hadoop

hadoop@hadoop:~$ ./jdk-6u15-linux-i586.bin

hadoop@hadoop:~$ ls
hypertable-0.9.2.6-Alpha-src.tar.gz  jdk1.6.0_15  jdk-6u15-linux-i586.bin
hadoop@hadoop:~$ tar xvzf hypertable-0.9.2.6-Alpha-src.tar.gz

Create an install directory (optional)

mkdir ~/hypertable

Create a build directory

mkdir -p ~/build/hypertable

hadoop@hadoop:~$ cd ~/build/hypertable
hadoop@hadoop:~/build/hypertable$
hadoop@hadoop:~/build/hypertable$ cmake ~/hypertable
hypertable/                          hypertable-0.9.2.5-Alpha-src.tar.gz
hypertable-0.9.2.5-Alpha/
hadoop@hadoop:~/build/hypertable$ cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=~/hypertable ~/ hypertable-0.9.2.5-Alpha

hadoop@hadoop:~/build/hypertable$ cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=~/hypertable ~/ hypertable-0.9.2.5-Alpha

hadoop@hadoop:~/build/hypertable$ make -j 3

這一步時間比較長，並且機器滿負載運行，可以喝杯咖啡在來看結果。

hadoop@hadoop:~/build/hypertable$ make install

接下來驗證安裝是否成功：

hadoop@hadoop:~/build/hypertable$  make alltests
Scanning dependencies of target runtestservers

Starting test servers


Shutdown hyperspace complete


ERROR: DfsBroker not running, database not cleaned


Shutdown thrift broker complete


Shutdown hypertable master complete


Shutdown range server complete


DFS broker: available file descriptors: 1024


Successfully started DFSBroker (local)


Successfully started Hyperspace


Successfully started Hypertable.Master


Successfully started Hypertable.RangeServer


Successfully started ThriftBroker


Built target runtestservers


Scanning dependencies of target alltests


Running tests...


Start processing tests


Test project /home/hadoop/build/hypertable


  1/ 60 Testing Common-Exception                 Passed


2/ 60 Testing Common-Logging                   Passed


  3/ 60 Testing Common-Serialization             Passed


  4/ 60 Testing Common-ScopeGuard                Passed


  5/ 60 Testing Common-InetAddr                  Passed


  6/ 60 Testing Common-PageArena                 Passed


  7/ 60 Testing Common-Properties                Passed


8/ 60 Testing Common-BloomFilter               Passed


  9/ 60 Testing Common-Hash                      Passed


10/ 60 Testing HyperComm                        Passed


11/ 60 Testing HyperComm-datagram               Passed


12/ 60 Testing HyperComm-timeout                Passed


13/ 60 Testing HyperComm-timer                  Passed


14/ 60 Testing HyperComm-reverse-request        Passed


15/ 60 Testing BerkeleyDbFilesystem             Passed


16/ 60 Testing FileBlockCache                   Passed


17/ 60 Testing TableIdCache                     Passed


18/ 60 Testing CellStoreScanner                 Passed


19/ 60 Testing CellStoreScanner-delete          Passed


20/ 60 Testing Schema                           Passed


21/ 60 Testing LocationCache                    Passed


22/ 60 Testing LoadDataSource                   Passed


23/ 60 Testing LoadDataEscape                   Passed


24/ 60 Testing BlockCompressor-BMZ              Passed


25/ 60 Testing BlockCompressor-LZO              Passed


26/ 60 Testing BlockCompressor-NONE             Passed


27/ 60 Testing BlockCompressor-QUICKLZ          Passed


28/ 60 Testing BlockCompressor-ZLIB             Passed


29/ 60 Testing CommitLog                        Passed


30/ 60 Testing MetaLog-RangeServer              Passed


31/ 60 Testing Client-large-block               Passed


32/ 60 Testing Client-periodic-flush            Passed


33/ 60 Testing HyperDfsBroker                   Passed


34/ 60 Testing Hyperspace                       Passed


35/ 60 Testing Hypertable-shell                 Passed


36/ 60 Testing Hypertable-shell-ldi-stdin       Passed


37/ 60 Testing RangeServer                      Passed


38/ 60 Testing ThriftClient-cpp                 Passed


39/ 60 Testing ThriftClient-perl                Passed


40/ 60 Testing ThriftClient-java                Passed


41/ 60 Testing Client-random-write-read         Passed


42/ 60 Testing RangeServer-commit-log-gc        Passed


43/ 60 Testing RangeServer-load-exception       Passed


44/ 60 Testing RangeServer-metadata-split       Passed


45/ 60 Testing RangeServer-maintenance-thread   Passed


46/ 60 Testing RangeServer-row-overflow         Passed


47/ 60 Testing RangeServer-rowkey-ag-imbalanc   Passed


48/ 60 Testing RangeServer-split-recovery-1     Passed


49/ 60 Testing RangeServer-split-recovery-2     Passed


50/ 60 Testing RangeServer-split-recovery-3     Passed


51/ 60 Testing RangeServer-split-recovery-4     Passed


52/ 60 Testing RangeServer-split-recovery-5     Passed


53/ 60 Testing RangeServer-split-recovery-6     Passed


54/ 60 Testing RangeServer-split-recovery-7     Passed


55/ 60 Testing RangeServer-split-recovery-8     Passed


56/ 60 Testing RangeServer-split-merge-loop10   Passed


57/ 60 Testing RangeServer-bloomfilter-rows     Passed


58/ 60 Testing RangeServer-bloomfilter-rows-c   Passed


59/ 60 Testing RangeServer-ScanLimit            Passed


60/ 60 Testing Client-no-log-sync               Passed

100% tests passed, 0 tests failed out of 60
Built target alltests
hadoop@hadoop:~/build/hypertable$

至此，hypertable單機版ok。

接著，我們來體驗一下架在hadoop的hdfs上的hypertable。

hadoop@hadoop:~/hadoop-0.20.0/conf$ more hadoop-env.sh
# Set Hadoop-specific environment variables here.

# The only required environment variable is JAVA_HOME.  All others are
# optional.  When running a distributed configuration it is best to
# set JAVA_HOME in this file, so that it is correctly defined on
# remote nodes.

# The java implementation to use.  Required.
export JAVA_HOME=~/jdk1.6.0_15

# Extra JAVA CLASSPATH elements.  Optional.
# export HADOOP_CLASSPATH=

# The maximum amount of heap to use, in MB. Default is 1000.
export HADOOP_HEAPSIZE=100

hadoop@hadoop:~/hadoop-0.20.0/conf$ vi core-site.xml
<?xml version="1.0"?>
<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>

<!-- Put site-specific property overrides in this file. -->

<configuration>
<property>
<name>hadoop.tmp.dir</name>
<value>/home/hadoop/hadoop-0.20.0/tmp/dir/hadoop-${user.name}</value>
<description>A base for other temporary directories.</description>
</property>

<property>
<name>fs.default.name</name>
<value>hdfs://localhost:9000</value>
<description>The name of the default file system.  A URI whose
scheme and authority determine the FileSystem implementation.  The
uri's scheme determines the config property (fs. SCHEME.impl) naming
the FileSystem implementation class.  The uri's authority is used to
determine the host, port, etc. for a filesystem.</description>
</property>

</configuration>

hadoop@hadoop:~/hadoop-0.20.0/conf$ vi mapred-site.xml
<?xml version="1.0"?>
<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>

<!-- Put site-specific property overrides in this file. -->

<configuration>
<property>
<name>mapred.job.tracker</name>
<value>localhost:9001</value>
<description>The host and port that the MapReduce job tracker runs
at.  If "local", then jobs are run in-process as a single map
and reduce task.
</description>
</property>

</configuration>
~

~
"mapred-site.

hadoop@hadoop:~/hadoop-0.20.0/conf$ vi hdfs-site.xml
<?xml version="1.0"?>
<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>

<!-- Put site-specific property overrides in this file. -->

<configuration>
<property>
<name>dfs.replication</name>
<value>1</value>
<description>Default block replication.
The actual number of replications can be specified when the file is created.
The default is used if replication is not specified in create time.
</description>
</property>

</configuration>
~
~

hadoop@hadoop:~$ more .bash_profile
export JAVA_HOME=~/jdk1.6.0_15
export HADOOP_HOME=~/hadoop-0.20.0
export PATH=$PATH:$JAVA_HOME/bin:$HADOOP_HOME/bin

hadoop@hadoop:~$ hadoop namenode -format
09/08/18 21:10:25 INFO namenode. NameNode: STARTUP_MSG:
/************************************************************
STARTUP_MSG: Starting NameNode
STARTUP_MSG:   host = hadoop/127.0.1.1
STARTUP_MSG:   args = [-format]
STARTUP _MSG:   version = 0.20.0
STARTUP_MSG:   build = HTTPs://svn.apache.org/repos/asf/hadoop/core/ branches/branch-0.20 -r 763504; compiled by 'ndaley' on Thu Apr  9 05:18:40 UTC 2009
********************************************************** **/
09/08/18 21:10:26 INFO namenode. FSNamesystem: fsOwner=hadoop,hadoop,adm,dialout,cdrom,plugdev,lpadmin,sambashare,admin
09/08/18 21:10:26 INFO namenode. FSNamesystem: supergroup=supergroup
09/08/18 21:10:26 INFO namenode. FSNamesystem: isPermissionEnabled=true
09/08/18 21:10:26 INFO common. Storage: Image file of size 96 saved in 0 seconds.
09/08/18 21:10:26 INFO common. Storage: Storage directory /home/hadoop/hadoop-0.20.0/tmp/dir/hadoop-hadoop/dfs/name has been successfully formatted.
09/08/18 21:10:26 INFO namenode. NameNode: SHUTDOWN_MSG:
/************************************************************
SHUTDOWN_MSG: Shutting down NameNode at hadoop/127.0.1.1
************************************************************/
hadoop@hadoop:~$

hadoop@hadoop:~$ start-all.sh
starting namenode, logging to /home/hadoop/hadoop-0.20.0/bin/.. /logs/hadoop-hadoop-namenode-hadoop.out
localhost: starting datanode, logging to /home/hadoop/hadoop-0.20.0/bin/.. /logs/hadoop-hadoop-datanode-hadoop.out
localhost: starting secondarynamenode, logging to /home/hadoop/hadoop-0.20.0/bin/.. /logs/hadoop-hadoop-secondarynamenode-hadoop.out
starting jobtracker, logging to /home/hadoop/hadoop-0.20.0/bin/.. /logs/hadoop-hadoop-jobtracker-hadoop.out
localhost: starting tasktracker, logging to /home/hadoop/hadoop-0.20.0/bin/.. /logs/hadoop-hadoop-tasktracker-hadoop.out

hadoop@hadoop:~$ jps
12959 JobTracker
12760 DataNode
12657 NameNode
13069 TaskTracker
13149 Jps
12876 SecondaryNameNode

ok，hadoop 0.20.0配置完成。

接著我們整合hypertable和hadoop hdfs。

hadoop@hadoop:~/hypertable/0.9.2.5/bin$ more ~/.bash_profile
export JAVA_HOME=~/jdk1.6.0_15
export HADOOP_HOME=~/hadoop-0.20.0
export HYPERTABLE_HOME=~/hypertable/0.9.2.5/
export PATH=$PATH:$JAVA_HOME/bin:$HADOOP_HOME/bin:$HYPERTABLE_HOME/bin

hadoop@hadoop:~/hypertable/0.9.2.5/conf$ ls
hypertable.cfg  METADATA.xml
hadoop@hadoop:~/hypertable/0.9.2.5/conf$ more hypertable.cfg
#
# hypertable.cfg
#

# Global properties
Hypertable.Request.Timeout=180000

# HDFS Broker
HdfsBroker.Port=38030
HdfsBroker.fs.default.name=hdfs://localhost:9000
HdfsBroker.Workers=20

# Local Broker
DfsBroker.Local.Port=38030
DfsBroker.Local.Root=fs/local

# DFS Broker - for clients
DfsBroker.Host=localhost
DfsBroker.Port=38030

# Hyperspace
Hyperspace.Master.Host=localhost
Hyperspace.Master.Port=38040
Hyperspace.Master.Dir=hyperspace
Hyperspace.Master.Workers=20

# Hypertable.Master
Hypertable.Master.Host=localhost
Hypertable.Master.Port=38050
Hypertable.Master.Workers=20

# Hypertable.RangeServer
Hypertable.RangeServer.Port=38060

Hyperspace.KeepAlive.Interval=30000
Hyperspace.Lease.Interval=1000000
Hyperspace.GracePeriod=200000

# ThriftBroker
ThriftBroker.Port=38080
hadoop@hadoop:~/hypertable/0.9.2.5/conf$

啟動基於hadoop的hypertalbe

hadoop@hadoop:~/hypertable/0.9.2.5/bin$ start-all-servers.sh hadoop
DFS broker: available file descriptors: 1024
Successfully started DFSBroker (hadoop)
Successfully started Hyperspace
Successfully started Hypertable.Master
Successfully started Hypertable.RangeServer
Successfully started ThriftBroker

hadoop@hadoop:~/hypertable/0.9.2.5/log$ hadoop fs -ls /
Found 2 items
drwxr-xr-x   - hadoop supergroup          0 2009-08-18 21:25 /home
drwxr-xr-x   - hadoop supergroup          0 2009-08-18 21:28 /hypertable
hadoop@hadoop:~/hypertable/0.9.2.5/log$ hadoop fs -ls /hypertable
Found 2 items
drwxr-xr-x   - hadoop supergroup          0 2009-08-18 21:28 /hypertable/server
drwxr-xr-x   - hadoop supergroup          0 2009-08-18 21:28 /hypertable/tables

至此，基於hadoop的hypertable成功啟動。 下一篇進行hql的體驗。