[autoscaler][aws/gcp/azure][job submission] Fix autoscaler defaults (r…

…ay-project#29741)

This fixes the default autoscaler configs so that the minimal examples are compatible (or more compatible with job submission). In particular, there are 2 fixes

Install ray[default] extras because the job submission portion of the api server depends on it.
Set ray start --head --dashboard-host=0.0.0.0 so that the dashboard listens for remote connections (instead of the default localhost).

Co-authored-by: Alex <[email protected]>

python/ray/autoscaler/aws/defaults.yaml

@@ -136,7 +136,7 @@ worker_setup_commands: []
 # Command to start ray on the head node. You don't need to change this.
 head_start_ray_commands:
     - ray stop
-    - ulimit -n 65536; ray start --head --port=6379 --object-manager-port=8076 --autoscaling-config=~/ray_bootstrap_config.yaml
+    - ulimit -n 65536; ray start --head --port=6379 --object-manager-port=8076 --autoscaling-config=~/ray_bootstrap_config.yaml --dashboard-host=0.0.0.0
 
 # Command to start ray on worker nodes. You don't need to change this.
 worker_start_ray_commands:

python/ray/autoscaler/azure/defaults.yaml

@@ -144,7 +144,7 @@ worker_setup_commands: []
 # Command to start ray on the head node. You don't need to change this.
 head_start_ray_commands:
     - ray stop
-    - ulimit -n 65536; ray start --head --port=6379 --object-manager-port=8076 --autoscaling-config=~/ray_bootstrap_config.yaml
+    - ulimit -n 65536; ray start --head --port=6379 --object-manager-port=8076 --autoscaling-config=~/ray_bootstrap_config.yaml --dashboard-host=0.0.0.0
 
 # Command to start ray on worker nodes. You don't need to change this.
 worker_start_ray_commands:

python/ray/autoscaler/gcp/defaults.yaml

@@ -139,7 +139,7 @@ setup_commands:
     - >-
         (stat /opt/conda/bin/ &> /dev/null &&
         echo 'export PATH="/opt/conda/bin:$PATH"' >> ~/.bashrc) || true
-    - which ray || pip install -U https://s3-us-west-2.amazonaws.com/ray-wheels/latest/ray-3.0.0.dev0-cp37-cp37m-manylinux2014_x86_64.whl
+    - which ray || pip install -U "ray[default] @ https://s3-us-west-2.amazonaws.com/ray-wheels/latest/ray-3.0.0.dev0-cp37-cp37m-manylinux2014_x86_64.whl"
 
 
 # Custom commands that will be run on the head node after common setup.
@@ -159,6 +159,7 @@ head_start_ray_commands:
       --port=6379
       --object-manager-port=8076
       --autoscaling-config=~/ray_bootstrap_config.yaml
+      --dashboard-host=0.0.0.0
 
 # Command to start ray on worker nodes. You don't need to change this.
 worker_start_ray_commands: