AiiDA on ubuntu 22.04
Ubuntu 22.04 installs recent RabbitMQ version that requires consumer_timeout configuration for AiiDA. See
RabbitMQ version to use · aiidateam/aiida-core Wiki · GitHub
What should be done is as follows:
sudo vi /etc/rabbitmq/rabbitmq.conf
add the line
consumer_timeout = 3600000000
Restart. RabbitMQ server
sudo systemctl restart rabbitmq-server.service
Check if this file is listed under the "Config files" section by
sudo rabbitmq-diagnostics status
If all OK, then
verdi config set warnings.rabbitmq_version false
AiiDA setup (aiida-core, aiida-vasp, aiida-phonoxpy) on an empty VM from scratch on Ubuntu 22.04
sudo apt install rabbitmq-server
Prepare miniforge and
conda install python postgresql
Installation into Conda environment — AiiDA 2.5.0.post0 documentation
git clone https://github.com/aiidateam/aiida-core.git cd aiida-core pip install -e ."[atomic_tools]"
Initialize postgres, and prepare aiidadb. Now verdi quicksetup works nicely.
initdb -D /home/username/aiida-proj pg_ctl -D /home/username/aiida-proj start verdi quicksetup
Tab completion may be set in $CONDA_PREFIX/etc/conda/activate.d/.
AiiDA plugins of aiida-vasp and aiida-phonoxpy are installed via github repositories with pip install -e . -vvv.
Specific scheduler plugin may be required for me.
Setting up computer and code.
How to run external codes — AiiDA 2.5.0.post0 documentation
For aiida-vasp POTCARs,
verdi data vasp-potcar uploadfamily ...
Maybe setting up local queueing system (gridengine), https://atztogo.hatenablog.com/entry/2020/08/07/105923
Files of calculation on AiiDA
Usage — AiiDA 2.5.0.post0 documentation
local_copy_list
Input files -> local file system -> remote working directory
remote_copy_list
Files on remote machine -> remote working directory
retrieve_list
Files on remote working directory -> local retrieved folder
retrieve_temporary_list
Files on remote working directory -> local temporary folder -> Files removed after parsing
AiiDA error handling
Dealing with Errors — AiiDA Tutorials
Migration of python version on conda for AiiDA
I install AiiDA on conda env. Python packages follow python release versions, and often newer python versions are preferable. But updating python version on a conda env by, e.g., conda update "python=3.11" fails when many packages were already installed because of the failure of solving dependency. Then, an easy choice is creating new env and install packages from scratch. This is much easier for me indeed.
Recent conda can rename env easily
conda rename -n orig_env backup_env
Create orig_env
conda create -n orig_env
Install packages (here python 3.11 for example)
conda install "python=3.11" postgresql
Having major version change of postgresql, db migration is necessary, see https://atztogo.hatenablog.com/entry/2023/05/06/154612.
Install other packages
conda install numpy scipy h5py pyyaml matplotlib-base c-compiler cxx-compiler "libblas=*=*mkl" spglib mkl-include boost eigen cmake ipython make
Activate (deactivate) setting has to be modified for zsh user like me.
cd $CONDA_PREFIX/etc/conda/activate.d/ mkdir bash mv * bash cp ~/.miniforge/envs/backup_env/etc/conda/activate.d/env_vars.sh .
Modify env_vars.sh. Do the same for deactivate. In my case following lines:
export AIIDA_PATH=~/aiida eval "$(_VERDI_COMPLETE=zsh_source verdi)" export HOST=`hostname` autoload bashcompinit bashcompinit emulate sh -c 'source $CONDA_PREFIX/etc/conda/activate.d/bash/activate-binutils_linux-64.sh' emulate sh -c 'source $CONDA_PREFIX/etc/conda/activate.d/bash/activate-gcc_linux-64.sh' emulate sh -c 'source $CONDA_PREFIX/etc/conda/activate.d/bash/activate-gxx_linux-64.sh' emulate sh -c 'source $CONDA_PREFIX/etc/conda/activate.d/bash/libblas_mkl_activate.sh' emulate sh -c 'source $CONDA_PREFIX/etc/conda/activate.d/bash/libxml2_activate.sh'
are written in env_vars.sh. Then installing packages via pip. git-pull aiida-core, and
pip install -e ."[atomic_tools,docs,tests,rest]"
git-pull aiida-vasp, and
pip install -e ."[pre-commit,tests,graphs]"
Installed a different version of aiida-core, aiida-storage update may be necessary, but this is informed when verdi status or verdi daemon start, etc. So just follow it.
Transferring data in AiiDA
How to work with data — AiiDA 2.5.0.post0 documentation
Upgrade conda-forge postgresql from 14 to 15
Assume having old postgresql (v14) db directory exists at
/home/username/aiida-proj/psql
Upgrade postgresql (14 -> 15) of the current conda env (myproj)
Create a new conda env (psql) with old postgresql (14).
initdb -D /home/username/myproj/psql15 cp /home/username/myproj/psql/pg_hba.conf /home/username/myproj/psql15/ cp /home/username/myproj/psql/postgresql.conf /home/username/myproj/psql15/ pg_upgrade --old-datadir=/home/username/myproj/psql --new-datadir=/home/username/myproj/psql15 --old-bindir=/home/username/.miniforge/envs/psql/bin --new-bindir=/home/username/.miniforge/envs/myproj/bin --old-options '-c config_file=/home/username/myproj/psql/postgresql.conf' --new-options '-c config_file=/home/username/myproj/psql15/postgresql.conf' --check
If all OK, remove --check and rerun the above again.
mv /home/username/myproj/psql /home/username/myproj/psql14 ln -s /home/username/myproj/psql15 /home/username/myproj/psql
AiiDA v2.0 migration
- How to manage your installation — AiiDA 2.0.0 documentation
- Changelog — AiiDA 2.5.0.post0 documentation
- Data types — AiiDA 2.5.0.post0 documentation
eval "$(_VERDI_COMPLETE=zsh_source verdi)" export AIIDA_WARN_v3=1
% conda update pip
In aiida-core
% find . -name "*.pyc" -delete % rm -r ./*.egg-info % git checkout develop % git pull % git checkout v2.0.1 % pip install -e ."[atomic_tools,docs,tests,rest]" % verdi storage migrate % verdi storage maintain --full
Migration log and old data
migration-repository-missing-subfolder-qxze5vt4.json .aiida/repository/demo/repository
