Using hbase as an example to apply database watermarking algorithm to big data

Compiler Environment

• Operating System : Centos7

• Hadoop : 2.6.0

• boost : 1.47.0

• libevent : 2.0.12-stable

• Hbase : 0.98.8

• Thrift : 0.7.0

• jdk : 1.8.0_221

Note

We use hbase in the Standalone mode.

How to run it

1. You should start related server.

start-dfs.sh

start-yarn.sh

/usr/local/hbase-0.98.8/bin/start-hbase.sh

/usr/local/hbase-0.98.8/bin/hbase-daemon.sh start thrift2

jps

4130 ResourceManager
3957 SecondaryNameNode
3639 NameNode
4922 HMaster
5309 Jps
5230 ThriftServer
4239 NodeManager

2. Create a table 'hbase_test' in the hbase.

/usr/local/hbase-0.98.8/bin/hbase shell create 'hbase_test', 'info' scan 'hbase_test' exit

3. Compile the data2hbase.cpp and run it to write data to database.

cd /home/hadoop

• g++ -DHAVE_NETINET_IN_H -o Data2HbaseClient -I/usr/local/thrift/include/thrift -I./gen-cpp -L/usr/local/thrift/lib data2hbase.cpp ./gen-cpp/hbase_types.cpp ./gen-cpp/hbase_constants.cpp ./gen-cpp/THBaseService.cpp -lthrift -g

• ./Data2HbaseClient 0.0.0.0 9090

If there is any warning, try

source /etc/profile

4. Compile the watermark.cpp and excute the watermarking algorithm.

• g++ -DHAVE_NETINET_IN_H -o WatermarkClient -I/usr/local/thrift/include/thrift -I./gen-cpp -L/usr/local/thrift/lib watermark.cpp SHA1.cpp ./gen-cpp/hbase_types.cpp ./gen-cpp/hbase_constants.cpp ./gen-cpp/THBaseService.cpp -lthrift -g

• ./WatermarkClient 0.0.0.0 9090