For testing, just build main.go for each command:
go build -o cmd/datasetIngestor/datasetIngestor cmd/datasetIngestor/main.go
To build all applications and target architectures, use the build script:
cmd/build.sh
Tools are compiled for the following architectures:
- linux / amd64
- windows / amd64
- macos / universal
These can be cross-compiled from any system.
Tools are deployed by committing the binaries to this repository.
-
Run
deploy.sh. This script- Builds executables for all targets
- Commits all changes to the git rollout repository.
After this step the version numbers are out of sync until the deploy steps below are performed.
-
Deploy to the scicat tools repo. This is the public deployment
- Following this step the commands can be fetched via curl from https://gitlab.psi.ch/scicat/tools/$OS
-
Deploy linux versions to online beamline consoles (you need to have write access rights):
cd linux
scp datasetArchiver datasetIngestor datasetRetriever datasetGetProposal datasetCleaner SciCat [email protected]:/work/sls/bin/
- Deploy linux versions to the ingest server pbaingest01
ssh [email protected]
cd bin/
curl -O https://gitlab.psi.ch/scicat/tools/raw/master/linux/datasetIngestor;chmod +x datasetIngestor
curl -O https://gitlab.psi.ch/scicat/tools/raw/master/linux/datasetArchiver;chmod +x datasetArchiver
curl -O https://gitlab.psi.ch/scicat/tools/raw/master/linux/datasetGetProposal;chmod +x datasetGetProposal
- Deploy to Ra cluster as a pmodule
cd pmodules/buildblocks/Tools/datacatalog
kinit
aklog
# optionally update files/variants.Linux to new version
# the following may be needed first:
#export PMODULES_TMPDIR=/var/tmp/egli
module load Pmodules/1.0.0rc13
./build 1.1.10 -f