Installation

Installing From Source

Installation from source requires several tools and external packages to be available as well as the building and installation of three Odin and Tristan specific software packages. This guide assumes the the installation is taking place on a CentOS 7 (or RHEL 7) operating system. The source can also be built for other operating systems; any operating system specific package installation commands should simply be replaced with the equivalent command for the chosen operating system.

Prerequisites

The Odin and Tristan software requires standard C, C++ and python development tools to be present. The following libraries and tools are also required:

All of the above packages can be installed from the default package manager Yum (https://www.redhat.com/sysadmin/how-manage-packages) with the following commands:

yum install -y epel-release
yum groupinstall -y 'Development Tools'
yum install -y boost boost-devel
yum install -y cmake
yum install -y log4cxx-devel
yum install -y zeromq-devel
yum install -y libpcap-devel
yum install -y python2-pip
yum install -y python-virtualenv
yum install -y Cython

For all remaining sections of this installation guide it is assumed that source files are going to be located in the directory:

/home/tristan/

This directory will be referred to as the $SRC directory:

export SRC=/home/tristan

Install HDF5 libraries

It is necessary to install the HDF5 libraries to use Odin Data to save the data to disk. There are prebuilt installation files available and also some OS distributions provide specific packages to install the HDF libraries. This method shows how HDF5 can be installed from source if preferred. Start by downloading and installing the szip library:

cd $SRC
wget https://support.hdfgroup.org/ftp/lib-external/szip/2.1.1/src/szip-2.1.1.tar.gz
gunzip szip-2.1.1.tar.gz
tar -xvf szip-2.1.1.tar
cd szip-2.1.1
./configure --prefix=$SRC/szip-2.1.1/install

Note this installs the library into a local relative path. This will allow a non-priveliged user to perform the installation if required. Now make and install the library:

make
make install

Now download, unpack and configure the main package:

cd $SRC
wget -O hdf5-1.12.0.tar.gz https://www.hdfgroup.org/package/hdf5-1-12-0-tar-gz/?wpdmdl=14582
gunzip hdf5-1.12.0.tar.gz
tar -xvf hdf5-1.12.0.tar
cd hdf5-1.12.0
./configure --with-szlib=$SRC/szip-2.1.1/install --prefix=$SRC/hdf5-1.12.0/install --enable-threadsafe --enable-hl --with-pthread --enable-shared --enable-unsupported

Once again the configuration line above configures for a local install. Now make and install:

make
make install

The installation of the HDF5 libraries should now be complete. There are various tools and tests that can be run to verify the installation. Visit https://www.hdfgroup.org/ for more details regarding the HDF5 libraries and tools.

Install Python Virtual Environment

The current version of Tristan control software runs on Python version 2.7. The easiest method of setting up and installing an environment is to use pip. Start by creating the virtual environment:

cd $SRC
virtualenv venv27

Now activate the environment for use:

source $SRC/venv27/bin/activate

Update the version of pip and the virtual environment:

pip install --upgrade pip
pip install --upgrade virtualenv

This virtual environment will be used to install all of the required modules for the detector software.

Installing Odin-Control

Odin-Control contains the control server which will act as the central point of control for all other parts of the software stack. This module contains only Python. Start by downloading the odin-control module:

cd $SRC
wget -O odin-control.zip https://github.com/dls-controls/odin-control/archive/external-build.zip
unzip odin-control.zip
mv odin-control-external-build/ odin-control

If preferred the module can be cloned from github using git:

cd $SRC
git clone https://github.com/dls-controls/odin-control.git
cd odin-control
git checkout external-build

Python Installation

Activate the Python virtual environment (if not already activated) and install the requirements and then the module itself:

source $SRC/venv27/bin/activate
cd $SRC/odin-control/
pip install -r requirements.txt
python ./setup.py install

At this stage it is possible to test that the installation was successful by requesting the command line help of the Odin control server application:

odin_server --help

Installing Odin-Data

Odin-Data contains both Python and C++ software. The main data acquisition base software is contained (FrameReceiver and FrameProcessor applications) as well as the necessary Python control software (Python adapters that are loaded into the Odin control server) to control the data acquisition applications. Finally, this module contains the application known as the MetaWriter which Tristan also requires. At the current time of writing a development branch is required to run the Tristan software.

To start the installation, download the odin-data module:

cd $SRC
wget -O odin-data.zip https://github.com/dls-controls/odin-data/archive/dev_branch.zip
unzip odin-data.zip
mv odin-data-dev_branch odin-data
cd odin-data

If preferred the module can be cloned from github using git:

cd $SRC
git clone https://github.com/dls-controls/odin-data.git
cd odin-data
git checkout dev_branch

Python Installation

Activate the Python virtual environment (if not already activated) and install the requirements and then the module itself:

source $SRC/venv27/bin/activate
cd $SRC/odin-data/tools/python
pip install -r requirements.txt
python ./setup.py install

At this stage it is possible to test that the Python installation was successful by requesting the command line help of the MetaWriter application:

meta_writer --help

C++ Installation

The C++ applications and tests are built using the CMake tool. Start by creating a build directory and configuring the module:

cd $SRC/odin-data/
mkdir build_dir
cd build_dir
cmake -DHDF5_ROOT=$SRC/hdf5-1.12.0/install -DCMAKE_INSTALL_RPATH_USE_LINK_PATH=ON -DCMAKE_INSTALL_PREFIX=$SRC/odin-data/install ..

Now make and then install the module:

make
make install

Once the installation has completed the unit tests can be executed from the current directory:

./bin/frameProcessorTest
./bin/frameReceiverTest

To verify the installation change to the install directory and run one of the applications with the help command line option:

cd $SRC/odin-data/install
./bin/frameReceiver --help

Installing Tristan-Detector

Tristan-Detector contains both Python and C++ software. This module contains the control and data acquisition libraries that are specific to the detector, and these are loaded into the control and data acquisition applications from the other modules.

To start the installation, download the odin-data module:

cd $SRC
wget -O tristan-detector.zip https://github.com/dls-controls/tristan-detector/archive/meta.zip
unzip tristan-detector.zip
mv tristan-detector-meta tristan-detector
cd tristan-detector

If preferred the module can be cloned from github using git:

cd $SRC
git clone https://github.com/dls-controls/tristan-detector.git
cd tristan-detector
git checkout meta

Python Installation

Activate the Python virtual environment (if not already activated) and install the requirements and then the module itself:

source $SRC/venv27/bin/activate
cd $SRC/tristan-detector/control
pip install -r control_requirements.txt
python ./setup.py install

At this stage it is possible to test that the Python installation was successful by requesting the command line help of the simulator application:

tristan_simulator --help

C++ Installation

The C++ applications and tests are built using the CMake tool. Start by creating a build directory and configuring the module:

cd $SRC/tristan-detector/
mkdir build_dir
cd build_dir
cmake -DODINDATA_ROOT_DIR=$SRC/odin-data/install -DCMAKE_INSTALL_RPATH_USE_LINK_PATH=ON -DCMAKE_INSTALL_PREFIX=$SRC/tristan-detector/install ..

Now make and then install the module:

make
make install

Once the installation has completed the unit tests can be executed from the current directory:

./bin/LATRDDecoderTest
./bin/LATRDProcessorTest

Next Steps

Once the installation is complete it is possible to create a simulated instance of the detector to verify all of the components are running correctly. The setting up guide will explain the steps necessary to achieve this.