doc/source/ray-core/starting-ray.rst
.. _start-ray:
This page covers how to start Ray on your single machine or cluster of machines.
.. tip:: Be sure to have :ref:installed Ray <installation> before following the instructions on this page.
Ray programs are able to parallelize and distribute by leveraging an underlying Ray runtime. The Ray runtime consists of multiple services/processes started in the background for communication, data transfer, scheduling, and more. The Ray runtime can be started on a laptop, a single server, or multiple servers.
There are three ways of starting the Ray runtime:
ray.init() (:ref:start-ray-init)start-ray-cli)start-ray-up)In all cases, ray.init() will try to automatically find a Ray instance to
connect to. It checks, in order:
RAY_ADDRESS OS environment variable.ray.init(address=<address>).ray start... _start-ray-init:
Calling ray.init() starts a local Ray instance on your laptop/machine. This laptop/machine becomes the "head node".
.. note::
In recent versions of Ray (>=1.5), ray.init() will automatically be called on the first use of a Ray remote API.
.. tab-set::
.. tab-item:: Python
.. testcode::
:hide:
import ray
ray.shutdown()
.. testcode::
import ray
# Other Ray APIs will not work until `ray.init()` is called.
ray.init()
.. tab-item:: Java
.. code-block:: java
import io.ray.api.Ray;
public class MyRayApp {
public static void main(String[] args) {
// Other Ray APIs will not work until `Ray.init()` is called.
Ray.init();
...
}
}
.. tab-item:: C++
.. code-block:: c++
#include <ray/api.h>
// Other Ray APIs will not work until `ray::Init()` is called.
ray::Init()
When the process calling ray.init() terminates, the Ray runtime will also terminate. To explicitly stop or restart Ray, use the shutdown API.
.. note::
The behavior of ray.shutdown() depends on how the cluster was initialized:
ray.init() started a new local cluster, ray.shutdown() will terminate all the local Ray processes.ray.init(address="auto") or ray.init(address="ray://<ip>:<port>")), ray.shutdown() only disconnects the client -- it does not shut down the remote cluster... tab-set::
.. tab-item:: Python
.. testcode::
:hide:
ray.shutdown()
.. testcode::
import ray
ray.init()
... # ray program
ray.shutdown()
.. tab-item:: Java
.. code-block:: java
import io.ray.api.Ray;
public class MyRayApp {
public static void main(String[] args) {
Ray.init();
... // ray program
Ray.shutdown();
}
}
.. tab-item:: C++
.. code-block:: c++
#include <ray/api.h>
ray::Init()
... // ray program
ray::Shutdown()
To check if Ray is initialized, use the is_initialized API.
.. tab-set::
.. tab-item:: Python
.. testcode::
import ray
ray.init()
assert ray.is_initialized()
ray.shutdown()
assert not ray.is_initialized()
.. tab-item:: Java
.. code-block:: java
import io.ray.api.Ray;
public class MyRayApp {
public static void main(String[] args) {
Ray.init();
Assert.assertTrue(Ray.isInitialized());
Ray.shutdown();
Assert.assertFalse(Ray.isInitialized());
}
}
.. tab-item:: C++
.. code-block:: c++
#include <ray/api.h>
int main(int argc, char **argv) {
ray::Init();
assert(ray::IsInitialized());
ray::Shutdown();
assert(!ray::IsInitialized());
}
See the Configuration <configure.html>__ documentation for the various ways to configure Ray.
.. _start-ray-cli:
ray start)Use ray start from the CLI to start a 1 node ray runtime on a machine. This machine becomes the "head node".
.. code-block:: bash
$ ray start --head --port=6379
Local node IP: 192.123.1.123 2020-09-20 10:38:54,193 INFO services.py:1166 -- View the Ray dashboard at http://localhost:8265
...
You can connect to this Ray instance by starting a driver process on the same node as where you ran ray start.
ray.init() will now automatically connect to the latest Ray instance.
.. tab-set::
.. tab-item:: Python
.. testcode::
import ray
ray.init()
.. tab-item:: java
.. code-block:: java
import io.ray.api.Ray;
public class MyRayApp {
public static void main(String[] args) {
Ray.init();
...
}
}
.. code-block:: bash
java -classpath <classpath> \
-Dray.address=<address> \
<classname> <args>
.. tab-item:: C++
.. code-block:: c++
#include <ray/api.h>
int main(int argc, char **argv) {
ray::Init();
...
}
.. code-block:: bash
RAY_ADDRESS=<address> ./<binary> <args>
You can connect other nodes to the head node, creating a Ray cluster by also calling ray start on those nodes. See :ref:on-prem for more details. Calling ray.init() on any of the cluster machines will connect to the same Ray cluster.
.. _start-ray-up:
ray up)Ray clusters can be launched with the :ref:Cluster Launcher <cluster-index>.
The ray up command uses the Ray cluster launcher to start a cluster on the cloud, creating a designated "head node" and worker nodes. Underneath the hood, it automatically calls ray start to create a Ray cluster.
Your code only needs to execute on one machine in the cluster (usually the head node). Read more about :ref:running programs on a Ray cluster <cluster-index>.
To connect to the Ray cluster, call ray.init from one of the machines in the cluster. This will connect to the latest Ray cluster:
.. testcode:: :hide:
ray.shutdown()
.. testcode::
ray.init()
Note that the machine calling ray up will not be considered as part of the Ray cluster, and therefore calling ray.init on that same machine will not attach to the cluster.
Check out our Deployment section <../cluster/getting-started.html>_ for more information on deploying Ray in different settings, including Kubernetes <../cluster/kubernetes/index.html>, YARN <../cluster/vms/user-guides/community/yarn.html>, and SLURM <../cluster/vms/user-guides/community/slurm.html>_.