Post-Installation Configuration and Startup

Depending on the method you chose to install AnzoGraph, there may be additional configuration steps you need to perform following installation. This topic describes steps you may need to perform following AnzoGraph installation, as well as other optional installation or configuration steps.

When the installer is finished, it creates an anzograph subdirectory in which it installs AnzoGraph files. So, for example, if you selected the default directory location, /opt/cambridgesemantics, the installer would create a subdirectory named /opt/cambridgesemantics/anzograph, where various AnzoGraph executable, configuration, example, and other program files would be installed. In addition, other AnzoGraph files such as the frontend, example systemd service and repo files, and the uninstall script are saved at the same level as the anzograph subdirectory.

Setting the JAVA_HOME Environmental Setting

Applicable only for non-root installations. JAVA_HOME is automatically set via a systemd service for Installer installations performed as root (sudo). Example systemd service files are created in the <Install_dir>/anzograph/examples subdirectory for non-root installations.

Installing Software Packages to Run AnzoGraph Extensions

Only required if you chose to install AnzoGraph UDXs (user-defined extensions). If you chose to install the AnzoGraph UDXs when running the Installer, after the AnzoGraph installation is completed, you will also need to install additional C++ dependency and support libraries required to run some of the UDXs. These files must be installed on every host server in which you installed AnzoGraph (standalone single server or each host server in a cluster).

Here is a list of the required software packages you need to install:

libarmadillo10
libboost_filesystem1_71_0
libboost_iostreams1_71_0
libboost_system1_71_0
libgrpc++1
libflatbuffers1
libhdfs3
libnfs13
libserd-0-0
shadow-utils

After AnzoGraph installation, you can use the .repo file located in the <Install_Dir>/examples/yum.repos.d directory to configure the yum repository on your system so that required UDX software packages can be installed.

After configuring the yum repository to install the UDX software packages, copy the csi-obs-cambridgesemantics-udxcontrib.repo file to /etc/yum.repos.d on your host server. You can then run the yum utility to install the required library files on each host server (standalone single server or each host server in a cluster). For example:

yum install --enablerepo=csi-obs-cambridgesemantics-udxcontrib libarchive libarmadillo10 libboost_filesystem1_71_0 libboost_iostreams1_71_0 libboost_system1_71_0 libgrpc++1 libflatbuffers1 libhdfs3 libnfs13 libserd-0-0 shadow-utils

The ability to write to the /etc/yum.repos.d directory requires root access permissions. See Adding, Enabling, and Disabling a YUM Repository for more information on defining and using yum repositories.

Post-Install GDI and Data Driver Configuration

The GDI and associated drivers are automatically installed in the anzograph/lib/udx directory by the Installer. No post installation configuration is required but the systemd services still need to be set up and the Frontend needs to be started, if you installed it. When running the Installer as root (sudo), the AnzoGraph systemd services are automatically configured and you can choose to automatically start services after the AnzoGraph installation has finished.

Cluster Leader and Compute Node IP Address Configuration

Only required for tarball installations. For Installer installations performed using root (sudo), the Installer prompts for entry of host servers in a cluster. For non-sudo installations (and using the archive TAR file), after installing AnzoGraph on each host server in a cluster environment, you need to configure the IP addresses and host server roles in the Install_dir/anzograph/config/ip_addrs.conf configuration file on each host server in the cluster. One host server is designated as the leader node and other servers in the cluster as compute nodes. (Typically users choose the lowest IP address to become the leader.)

  1. On each server in the cluster, open the Install_dir/anzograph/config/ip_addrs.conf file in a text editor.
  2. In the ip_addrs.conf file, enter the list of the IP addresses for each server in the cluster. Include each address on a new line. Type the leader server IP address on the first line, followed by each compute/worker server IP address. For example, the following entries define a cluster with four servers with the host server 192.168.2.1 designated as the leader server:
    192.168.2.1
    192.168.2.2
    192.168.2.3
    192.168.2.4

    Make sure that the ip_addrs.conf file is configured exactly the same, with IP addresses in the same order, on each server.

  3. Save and close the file.

Setting up AnzoGraph Services

When running the Installer as root (sudo), the Installer automatically creates systemd services in the /etc/systemd/system directory for azgmgrd, anzograph, and for jetty (if the AnzoGraph is installed). In addition, the Installer prompts whether you want to automatically start the services, after the Installation is finished.

When running the Installer as a non-root user, the Installer does not automatically create services, but example systemd service files are available in the <Install_dir>/examples/systemd-services directory for you to customize and start manually. After tailoring settings and various configuration options in the service files to your specific environment, copy the azgmgrd service file to the /etc/systemd/system directory on each host server (standalone single server or each host server in a cluster). The anzograph and jetty services are only installed on the host server in a standalone single server environment or on the leader node in cluster deployment of AnzoGraph.

For additional information about AnzoGraph services, see Starting and Stopping AnzoGraph.

Starting and Running AnzoGraph

For tarball installations only. To set up communication between host servers in a cluster you need to run the system manager daemon, a lightweight program that runs at all times and manages the communication between AnzoGraph servers. To do that:

  1. On each server in the cluster, run the following command to start the system manager daemon.
    ./install_path/bin/azgmgrd
  2. With the azgmgrd system management daemon running, on the leader node, run the following command to start AnzoGraph on the cluster:
    ./install_path/bin/azgctl -start

    For more information on starting and stopping AnzoGraph and performing other AnzoGraph graph database management operations, see Starting and Stopping AnzoGraph.

    With tarball file installations, the first time you start AnzoGraph you are prompted to review and accept the license agreement.

Where to Go From Here

For next steps on using AnzoGraph, see Quickstart Using the Command Line Interface, a brief tutorial designed to introduce you to the AnzoGraph command line interface and provide instructions for loading and querying a sample data set as well as your own data.

Related Topics