0 - Installation
Download the binary
Gaia Worker is embedded into the normal Gaia binary file. Therefore, the first step is to copy the Gaia binary file to your remote worker host or pull the Gaia docker image from docker hub.
Configure the worker
Before we can start the worker, we need to configure it. Since Gaia Worker uses the configuration loading mechanism like Gaia, you have a few options to provide the needed configuration.
Provide configuration via a configuration file. Create a new file called
.gaia_config in the same folder where the
Gaia binary is located with the following example configuration:
worker-host-url=http://gaia-url.com mode=worker worker-name=my-worker worker-grpc-host-url=gaia-url.com:8989 worker-tags=tag1,tag2,-python worker-secret=55d79ff1-37b9-595d-a973-65d50e95db85
You have the possibility to provide configuration via environment variables. See the example configuration below:
export GAIA_HOST_URL=http://localhost:8080 export GAIA_MODE=worker export GAIA_WORKER_NAME=my-worker export GAIA_WORKER_GRPC_HOST_URL=localhost:8989 export GAIA_WORKER_TAGS="tag1,tag2,-python" export GAIA_WORKER_SECRET=55d79ff1-37b9-595d-a973-65d50e95db85
- WORKER_HOST_URL (required) - The URL to the Gaia primary instance where this worker should be connected to.
- MODE (required) - Tells Gaia on start which mode it should start in. Since we want to start a worker it should be always “worker”.
- WORKE_RNAME (optional) - Gives the worker a name which is displayed in the primary Gaia UI. If empty, a random generated name will be set.
- WORKER_GRPC_HOSTURL (required) - The URL including the port to the Gaia primary instance. Should be mostly like
HOSTURLbut the port should be different.
- WORKER_TAGS (optional) - Tags workers which are matched during pipeline run schedule. Negative tags (for example: -python) can be used to deactivate automatic set tags by the worker.
- WORKER_SECRET (required) - The global registration secret which is used to register at the Gaia primary instance. Can be obtained in the primary Gaia instance UI (Settings -> Worker).
Starting the worker
As soon as the worker configuration is finished, we are able to start the worker. It should only take a few seconds until the worker is visible in the primary instance’s UI. If not, please have a look at the worker output logs.