OPC-UA extractor using libraries from OPCFoundation see here
Install using dotnet tool restore. Then run dotnet restore to install packages.
Now you can build projects using dotnet build [Project]. Using just dotnet build is likely to fail on "OpcUaServiceManager", since it cannot be built
with the dotnet CLI.
Install .NET 8.0 from here.
To run:
dotnet run --project ExtractorLauncher/.
There are a few environment variables that may be used:
- OPCUA_CONFIG_DIR
- OPCUA_CERTIFICATE_DIR
Beyond this, the .yml config file also supports environment variable substitution like ${SOME_ENVIRONMENT_VARIABLE}.
For CONFIG_DIR, default is [application dir]/config. CERTIFICATE_DIR is used for opc-ua certificates, and defaults to [application dir]/certificates.
See the example configuration for a config template.
There are docker images of each release at eu.gcr.io/cognite-registry/opcua-extractor-net.
Config, both opcua config opc.ua.extractor.Config.xml and config.yml are located in a volume /config and certificates are located in subfolders of the volume /certificates. Example:
docker run -v "$(pwd)/config:/config" -v "$(pwd)/certificates:/certificates eu.gcr.io/cognite-registry/opcua-extractor-net:tag
which would run the build tagged with tag using config stored in current_dir/config.
There will be binaries for the most recent release here on github. There are two different builds:
- windows x64, which should work for newer versions of windows (the 81 version will probably work for newer versions as well)
- linux x64, which should work on linux and osx.
There should be a system specific executable at the top level, which launches the extractor. These should be able to run with no extra setup.
In addition there are .deb, .rpm and .msi installers.
The extractor also takes a few command line arguments for convenience, use OpcuaExtractor -h for documentation.
Documentation can be found here
If you do not have access to an OPC-UA server, and wish to experiment with the extractor, you can run the test server, which is found in releases along with the version of the extractor it was used to test.
You can also build and run it by following the instructions in "Setup" above, then running dotnet run --project Server. Use dotnet run --project Server -- [options] to add command line options when running it this way.
The server has a command line interface, and is capable of simulating a lot of different server behavior. Run Server -h for documentation.
By default it runs on opc.tcp://localhost:62546, so that is where you would connect with UAExpert or the extractor.
The server is capable of generating history for events and simple and complex datapoints, as well as generating periodic updates, events and changes over time. It can also optionally simulate some types of buggy server behavior.
You will need .NET 8.0. Then simply run dotnet build to compile,
or dotnet run --project ExtractorLauncher to compile and run.
The compiler may complain about OpcUaExtractorSetup and OpcUaServiceManager, which aren't generally necessary to compile during development.
You can use dotnet build Test or dotnet build ExtractorLauncher to only compile some parts.
For testing metrics, a good solution is the prom-stack found here
To run the tests locally, run dotnet test, or use the test.sh script.
Some tests require an instance of influxdb 1.8 found here running on port 8086, and some require a version of mosquitto, found here running on port 4060. The tests now run their own OPC-UA servers.
During development it is perfectly fine to just run a subset of the tests using dotnet test --filter MyTestMethod. All tests run as part of CI.
Update manifest.yaml with the new release`. Merging to master will automatically
deploy binaries to github releases and docker images to eu.gcr.io/cognite-registry/. For
the MSI to build, the version must be simple, i.e. on the form "a.b.c".
Please follow the code style of the rest of the code, meaning:
- Public and protected members are properties and Capitalized.
- Private members are not prefixed, and should be camelCase.
- Curly braces on the next line.
In general just make sure the code remains consistent. The code has nullable analysis enabled, so make sure to fix any warnings. You should add tests for any new behavior or fixes.
This project adheres to Contributor Covenant v2.0 as a code of conduct.