The current version of NP-TMcode (M8.03) has been bundled in a pre-built `docker` image, distributed via `https://hub.docker.com/`. To use this image, run the command:
The current version of NP-TMcode (M10a) has been bundled in a pre-built `docker` image, distributed via `https://hub.docker.com/`. To use this image, run the command:
> docker run -it gmulas/np-tmcode-run:M8 /bin/bash
> docker run -it gmulas/np-tmcode-run:m10a.00 /bin/bash
The image will be automatically downloaded, if it is not available in the system, then it will be run in interactive mode, starting a shell session. Follow the instructions provided in Sec. 3.4.1 of the Release Notes document to proceed with testing and code use.
The image will be automatically downloaded, if it is not available in the system, then it will be run in interactive mode, starting a shell session. Follow the instructions provided in Sec. 3.4.1 of the Release Notes document to proceed with testing and code use.
The current version of NP-TMcode (M8.03) has been bundled in a pre-built `singularity` image, distributed at `https://www.ict.inaf.it/gitlab/giacomo.mulas/np-tmcode-singularity/-/releases/np-tmcode-singularity-run-M8.03`. To use this image, download and unpack the pre-built image bundle, then place the `np-tmcode-run.sif` file in the `containers/singularity` folder of the NP-TMcode project.
The current version of NP-TMcode (M10a) has been bundled in a pre-built `singularity` image, available from the sylabs cloud repository of singularity images at `https://cloud.sylabs.io/library/gmulas/np-tmcode-run/np-tmcode-run.sif`. To use this image, pull it with the command:
then place the `np-tmcode-run.sif` file in the `containers/singularity` folder of the NP-TMcode project.
## Building a local image
## Building a local image
A local image containing the necessary code requirements and binary files can be built through the image definition file provided along with the code. This `np-tmcode-run.def` file allows to create a `np-tmcode-run` image, that contains only the pre-built executables, python test scripts, compiled documentation, and the minimal runtime to run them.
A local image containing the necessary code requirements and binary files can be built through the image definition file provided along with the code. This `np-tmcode-run.def` file allows to create a `np-tmcode-run` image, that contains only the pre-built executables, python test scripts, compiled documentation, and the minimal runtime to run them.
...
@@ -20,7 +24,8 @@ or directly as
...
@@ -20,7 +24,8 @@ or directly as
> \<full path to image name\> \<np-tmcode command\>
> \<full path to image name\> \<np-tmcode command\>
where `<full path to image name>` is the name of the sif image, including full or relative path to it, and `<np-tmcode command>` is one of the executables built in np-tmcode project, or the `pycompare.py` script used to compare results between the Fortran and C++ versions of the code. These executables, as well as the minimum runtime for them to execute correctly, are embedded in the image, and "see" the current directory the image is run in (which technically is automatically bind-mounted by singularity in the running image instance).
where `<full path to image name>` is the name of the sif image, including full or relative path to it, and `<np-tmcode command>` is one of the executables built in np-tmcode project, or one of the accessory python scripts distributed with the code. These executables, as well as the minimum runtime for them to execute correctly, are embedded in the image, and "see" the current directory the image is run in (which technically is automatically bind-mounted by singularity in the running image instance). To further ease their use, one can also use the provided wrapper shell script, by setting the `SIFFILE` environment variable to the full path to the singularity image and then running the links with their respective names to `singwrapper` in the `scripts` directory.
