You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
By using our current JupyterLab image in the Dev cluster, let's eval and validate VScode as a viable IDE to develop MicroSim models and its integration with OpenM++.
It would be useful to get an evaluation of the default notebook resources, currently the CPU request is 0.5 but can scale to 4 cores, while the memory starts at 2GB and can scale to 4GB. Is this enough for basic model development and would additional resources (on the bursting side primarily) be an advantage?
The text was updated successfully, but these errors were encountered:
Models run, using both the UI and via the command line.
Models cannot be built due to a OpenM++ header file not being found.
libopenm folder not on PATH
libopenm is located at /opt/openm/1.15.4/openm
A ompp specific env variable called OM_ROOT should be set by the image to be /opt/openm/1.15.4/
export OM_ROOT=/opt/openmpp/1.15.4
The models build after setting this env variable.
Not sure things are being put in the right place though.
The env var should be set by the image and not by the user, but will work as a work-around until this is fixed.
There are other openm specific env vars used in the makefile that will need to be set.
OUT_PREFIX
BUILD_DIR
OM_MESSAGE_USE - used for mpi
MODEL_NAME - defaults to current dir if not set
RELEASE
OM_BD_LIB - location of sqlite
The OUT_PREFIX and BUILDDIR are user specific, meaning they depend on the path where the user put the model to be compiled, making this tricky to pre-set.
File associations have to be set in VSCode.
i.e. associate *.mpp and *.ompp files with *.cpp files
Can be set in Setting (gear icon)- Text Editor - Files - Associations
The openm wiki mentiones editing a setting.json file, but this does not seem to be an option in the installed version of VSCode, instead there being a wizard.
The mpp files are found and can be opened with VSCode, so this sees to be an option for syntax highlighting.
Epic: #2
By using our current JupyterLab image in the Dev cluster, let's eval and validate VScode as a viable IDE to develop MicroSim models and its integration with OpenM++.
The latest AAW jupyterlab-cpu image in dev/prod now includes c++ debugger and c++ extension for the built in vs code, tested following this guide: https://code.visualstudio.com/docs/cpp/config-msvc
Guide for getting started with model development: https://openmpp.org/wiki/openmpp-wiki.html#Linux-Create-and-Debug-Model
It would be useful to get an evaluation of the default notebook resources, currently the CPU request is 0.5 but can scale to 4 cores, while the memory starts at 2GB and can scale to 4GB. Is this enough for basic model development and would additional resources (on the bursting side primarily) be an advantage?
The text was updated successfully, but these errors were encountered: