The minimum platform requirements to run AI Reviewer are: 4GB of total system memory of which at least 2GB must be available to the JVM, and a Java Runtime Environment (JRE) v1.8 or greater (either 32 or 64 bit, depending on the specific version of AI Reviewer that you download).
However, we recommended a machine that has 8GB of total system memory or more, and a 64 bit version of the OS and JRE edition. Due to Win32 and 32 bit JRE limitations, the 32 bit version of AI Reviewer for Windows is restricted to a total of 1GB of memory, which may be insufficient for loading large projects. If you need to analyze large projects, please use Windows 64-bit or another platform.
First, download the binary package that is appropriate for your platform from the download section of the web site.
AI Reviewer follows the
unpack and run philosophy, so there is no dedicated installation program or script. Simply unpack the contents of the downloaded archive into a location of your choice (henceforth referred to as
the installation folder). Please note that current version of AI Reviewer does not support shared installation across multiple machines. In other words, you cannot run AI Reviewer on multiple machines from a single network shared deployment, although you may run multiple instances of AI Reviewer on the same machine.
By default, AI Reviewer is configured to utilize up to 2 GB of system memory. This is appropriate for most machines that have 4G of RAM. Depending on your machine’s configuration and the size of the projects you want to analyze, you may need to increase or decrease that limit by adjusting the parameter -Xmx in the configuration file ailauncher.ini, which for Linux and Windows systems is located in the root of the installation folder, and for MacOS systems is located under ailauncher.app/Contents/MacOS, inside the installation folder.
The argument -Xmx is one that tells the Java VM to allow the program (AI Reviewer) to ‘grow’ up to the given memory size. When analyzing multi-MLOC projects, AI Reviewer may require more RAM. In such situations, adjust the value of -Xmx as needed (e.g. -Xmx4096m will allow AI Reviewer to consume up to 4 GB of RAM).
At this stage, we recommend that you add the installation folder to the PATH environment variable, for convenience. Throughout the rest of the documentation, we assume that AI Reviewer’s installation folder is in the PATH. In order to test whether AI Reviewer is correctly installed, type the following command:
This should display usage information about the available command arguments of the tool, along with version and license information.
Every deployment of AI Reviewer must be activated, using the activation key supplied upon purchase. Depending on the type of license acquired, one and the same key may activate several deployments. Activation must be performed by a system administrator, or a user that has write permissions to the installation folder.
In order to activate a deployment, you need to run the following command
$ aireviewer -a
As a result, AI Reviewer will first check for the availability of an Internet connection. If an Internet connection is available, activation key and registration information will be requested from the user, and activation will be performed online. Please note that users accessing the Internet using a proxy server must configure proxy settings in the configuration file aireviewer.conf, located in the root of the installation folder.
If no Internet connection is possible, the user can opt to activate off-line. Off-line activation entails entering the same key and registration information as before, but the exchange of information with Logarix occurs via e-mail. In other words, the tool generates an activation request file (a text file of the form *.crequest), which the user is asked to send to Logarix support. In reply, Logarix support will send the user a license file (of the form *.license), which the user must subsequently import into AI Reviewer using the command:
$ aireviewer -a license_file
There are situations in which users may need to repeat activation for a pre-existing deployment, for example when renewing a subscription-based license that has or is about to expire. This is achieved using the exact same command as for a new activation:
$ aireviewer -a
This time however, the process will be fully automatic (provided that Internet connectivity is present), since AI Reviewer will not have to ask for the activation key and registration information to be typed in.