Installing OmniMV on CentOS

If an error is reported during Anaconda installation due to system compatibility issues, perform the following steps to install the dependencies.

1. Mount the CentOS image and modify the Yum source.

   mount CentOS-7-aarch64-Everything-1810.iso /media/

2. Create a Yum configuration file.
   1. Create an /etc/yum.repos.d/CentOS-local.repo file.

      vim /etc/yum.repos.d/CentOS-local.repo

   2. Press i to enter the insert mode and add the following content to the file:

      [local]
      name=CentOS-7.6 local
      baseurl=file:///media/
      enabled=1
      gpgcheck=0

   3. Press Esc, type :wq!, and press Enter to save the file and exit.
3. Make the configuration file take effect.

   yum clean all
   yum makecache

4. Install Python 3.10.2.
   1. Install the dependencies.

      yum install zlib-devel bzip2-devel openssl-devel ncurses-devel sqlite-devel readline-devel tk-devel gcc make libffi-devel

   2. Go to the /home directory and download the Python 3.10.2 source package.

      cd /home
      wget https://www.python.org/ftp/python/3.10.2/Python-3.10.2.tgz --no-check-certificate

   3. Decompress the Python source package and go to the /Python-3.10.2 directory.

      tar zxvf Python-3.10.2.tgz
      cd Python-3.10.2

   4. Configure the installation path.

      ./configure prefix=/usr/local/python3

   5. Perform compilation and installation.

      make && make install

   6. Modify the Python soft links.

      ln -s /usr/local/python3/bin/python3.10 /usr/bin/python3
      ln -s /usr/local/python3/bin/pip3.10 /usr/bin/pip3

5. Install the Python dependencies.
   • Spark engine dependency packages

     pip install Cython==0.29.24
     pip install numpy==1.22.3
     pip install pandas==1.3.5
     pip install PyYAML==6.0
     pip install mo-sql-parsing
     pip install sql-metadata==2.6.0
     pip install scikit-learn==1.1.2
     pip install xgboost==1.7.3
     pip install pyspark==3.1.1
     pip install pydoop==2.0.0

   • ClickHouse engine dependency packages

     pip install Cython==0.29.24
     pip install numpy==1.22.3
     pip install pandas==1.3.5
     pip install PyYAML==6.0
     pip install mo-sql-parsing
     pip install sql-metadata==2.6.0
     pip install clickhouse-driver==0.2.4
     pip install scikit-learn==1.1.2
     pip install xgboost==1.7.3

6. Download the software package of the materialized view recommendation algorithm.
   • Spark engine: Download the materialized view recommendation algorithm software package, and save the JAR package of the Spark SQL plugin and log parser JAR package to the server node of the Spark cluster. For details about how to obtain the software packages, see Obtaining Software. There is no special requirement on the software package paths. For example, you can store them in /opt/omnimv.

     mkdir -p /opt/omnimv
     mv boostkit-omnimv-spark-3.1.1-1.1.0-aarch64.jar /opt/omnimv
     mv boostkit-omnimv-logparser-spark-3.1.1-1.1.0-aarch64.jar /opt/omnimv
     mv BoostKit-omnimv_1.1.0.zip /opt/omnimv

   • ClickHouse engine:

     mkdir -p /opt/omnimv
     mv BoostKit-omnimv_1.1.0.zip /opt/omnimv

7. Decompress the OmniMV component package and remove the read and execute permissions of other user groups on the generated folder.

   cd /opt/omnimv
   unzip BoostKit-omnimv_1.1.0.zip
   chmod -R o-r BoostKit-omnimv_1.1.0

   

   You are advised to decompress the package as a non-root user so that the generated directory has the minimum permission to prevent files from being replaced.