Installing
Components can be easily added by adding a new top level element to your harperdb-config.yaml
file.
The configuration comprises two values:
component name - can be anything, as long as it follows valid YAML syntax.
package - a reference to your component.
Under the hood HarperDB is calling npm install
on all components, this means that the package value can be any valid npm reference such as a GitHub repo, an npm package, a tarball, a local directory, or a website.
When HarperDB is run or restarted it checks to see if there are any new or updated components. If there are, it will dynamically create a package.json file in the rootPath
directory and call npm install
.
npm will install all the components and store them in <ROOTPATH>/components
. A symlink back to <ROOTPATH>/node_modules
is also created for dependency resolution purposes.
The package.json file that is created will look something like this.
The package prefix is automatically added, however you can manually set it in your package reference.
Installing components using the operations API
To add a component using the operations API use the deploy_component
operation.
Another option is to pass deploy_component
a base64-encoded string representation of your component as a .tar
file. HarperDB can generate this via the package_component
operation. When deploying with a payload, your component will be deployed to your <ROOTPATH>/components
directory. Any components in this directory will be automatically picked up by HarperDB.
Last updated