我的书签
添加书签
移除书签Remote Pipeline Engine
Hop Remote pipeline engine
The remote run configuration runs pipelines on a remote Hop Server.
This run configuration requires little configuration, but has two prerequisites:
a running Hop Server
a Hop Server metadata definition in your project, pointing to your running Hop Server.
Options
| Option | Description |
|---|---|
Hop server | The remote Hop server to run this pipeline on. |
Run Configuration | The configuration to use on the remote Hop Server. In a lot of configurations, this will be a local (locally on the server) runtime configuration, but could be any other runtime configuration, e.g. a Spark runtime configuration that is started from a Spark cluster edge node |
Server poll delay (ms) | The delay in milliseconds between periodic polls to the remote server. If you don’t specify a value this defaults to 1000ms (1 second) |
Server poll interval (ms) | The interval in milliseconds between periodic polls to the remote server. If you don’t specify a value this defaults to 2000ms (2 seconds) |
Export linked resources to server? | Enable this option if you want to not only send the current pipeline to the server but also the referenced workflows and pipelines. See below for detailed information |
Named resources reference source folder | This is the reference source folder for the named resources that are being used (e.g. |
Named resources reference target folder | This is where you expect the source folder to map to on the server (e.g. |
Export linked resources to server
This option sends not just the parent pipeline to the server but also all other referenced pipelines and workflows. For example if you have mapping pipelines or workflows referenced they will be sent to the server as well. All the used pipelines and workflows together with the XML presentation of the pipeline execution configuration will be sent over to the server in the form of a ZIP archive. The server receives this archive and without unzipping runs the pipeline. To make this function correctly, Hop changes the references as well as references to filenames. For example ${PROJECT_HOME}/mapping.hpl will be changed to ${Internal.Entry.Current.Folder}/mapping.hpl. This means that it will try to use a relative path to the parent file.
If you have are using data files then those file names will be renamed as well. For example, you might be reading a file called ${PROJECT_HOME}/files/bigfile.csv in a CSV File Input transform. During the export the referenced filename will be changed to ${DATA_PATH_1}/bigfile.csv. For every folder that is referenced a new variable will be defined and set in the execution configuration. By default, the path set for this variable will be the same as on the executing (local) machine. On the server this might not make a lot of sense. For this reason you can specify a reference source folder like ${PROJECT_HOME} in combination with a target folder like /server/. In that example variable DATA_PATH_1 will get value /server/files/. This in turn allows you to transfer required files in advance or map a folder into a docker container and so on. It gives you flexibility when executing remotely while having ease of development on your client.
