Deployment at Point 5

The Point 5 machines do not have access to non-CERN resources, meaning that you cannot access GitHub or PyPI. For this reason, you will have to download a DQMGUI installation package and transfer it to your Point 5 machine.

1. Download an installation package

From a computer outside P5, download an installation package from the Releases page.

Each of the releases contains an installation package under Assets (look for the deployment...tar.gz file). Select the release that you prefer, and download it to a machine which can access GitHub (e.g. your personal machine you're reading this page from!).

You can also download the release with curl, e.g. for this release:

curl -L https://github.com/cms-DQM/dqmgui_prod_deployment/releases/download/python_3.8_deployment_HG2401b_dqmgui_10.0.0_root_v6-28-10/dqmgui_python_3.8_deployment_HG2401b_dqmgui_10.0.0_root_v6-28-10.tar.gz -O

2. Transfer the installation package to the P5 machine

From the machine where you downloaded the installation package, run scp to transfer the package:

scp -J <point 5 username>@cmsusr.cern.ch dqmgui_python_3.8_deployment_HG2401b_dqmgui_10.0.0_root_v6-28-10.tar.gz <point 5 username>@<point 5 machine name>:/tmp

3. Connect to the P5 machine and extract the package

From within the machine where you transferred the installation package to, run:

sudo su <your pro user>
cd ~
mkdir -p dqmgui_deployment
tar -xf /tmp/dqmgui_python_3.8_deployment_HG2401b_dqmgui_10.0.0_root_v6-28-10.tar.gz -C dqmgui_deployment

4. Start the installation

mkdir -p /data/srv
bash dqmgui_deployment/deploy_dqmgui.sh do_check_dependencies=0

The default installation directory is /data/srv. In case you are sharing the machine you are deploying to, it is advisable to modify it by specifying the INSTALLATION_DIR argument to the installation script, like so:

mkdir -p /data/ecal/srv
bash dqmgui_deployment/deploy_dqmgui.sh do_check_dependencies=0 INSTALLATION_DIR=/data/ecal/srv

The installation might take a considerable amount of time, depending on the number of cores available to the machine you're deploying on.