Skip to main content
Version: 1.x

Deploy and run

Seatunnel depends on Java operating environment and Spark. For detailed seatunnel installation steps, please refer to Installing seatunnel

The following highlights how the different platforms work:

Run seatunnel locally in local mode

./bin/ --master local[4] --deploy-mode client --config ./config/application.conf

Running seatunnel on Spark Standalone cluster

# client mode
./bin/ --master spark:// --deploy-mode client --config ./config/application.conf

# cluster mode
./bin/ --master spark:// --deploy-mode cluster --config ./config/application.conf

Running seatunnel on Yarn cluster

# client mode
./bin/ --master yarn --deploy-mode client --config ./config/application.conf

# cluster mode
./bin/ --master yarn --deploy-mode cluster --config ./config/application.conf

Running seatunnel on Mesos

# cluster mode
./bin/ --master mesos:// --deploy-mode cluster --config ./config/application.conf

The master, deploy-mode parameters of have the same meaning as Spark master, deploy-mode, Reference: Spark Submitting Applications

If you want to specify the resource size occupied by seatunnel when running, or other Spark parameters, you can specify it in the configuration file specified by --config:

spark {
spark.executor.instances = 2
spark.executor.cores = 1
spark.executor.memory = "1g"

For how to configure seatunnel, see seatunnel configuration