Release playground iamge and update doc

dev.md

@@ -49,9 +49,14 @@ According to this [tutorial](https://www.techrepublic.com/article/how-to-import-
 
 To run SQLFlow on a desktop computer running Windows, Linux, or macOS, you need to download
 
-1. the released `SQLFlowPlayground.ova`, directly download from [here](http://cdn.sqlflow.tech/latest/SQLFlowPlayground.ova), or use wget:
+1. the released VirtualBox `.ova` file, you have two choices:
+    - the minimized image (about 600M): shipped with all bootstrap files but no dependency docker images. When you start the playground, you will wait for a while to download the latest docker images, minikube framework and other packages.
     ```bash
-    wget -c http://cdn.sqlflow.tech/latest/SQLFlowPlayground.ova
+    wget -c http://cdn.sqlflow.tech/latest/SQLFlowPlaygroundBare.ova
+    ```
+    - the full installed image (about 2G): with all dependencies, no extra downloading is needed when starting. Note that in this case, the images will not be updated automatically, you will do it manually when needed.
+    ```bash
+    wget -c http://cdn.sqlflow.tech/latest/SQLFlowPlaygroundFull.ova
     ```
 1. optional, the [sqlflow](https://github.com/sql-machine-learning/sqlflow/blob/develop/doc/run/cli.md) command-line tool released by SQLFlow CI.
 

provision.bash

@@ -61,6 +61,7 @@ if which minikube > /dev/null; then
     echo "minikube installed. Skip."
 else
     $VAGRANT_SHARED_FOLDER/sqlflow/scripts/travis/install_minikube.sh
+    minikube config set WantUpdateNotification false
 fi
 echo "Done."
 

start.bash

@@ -29,14 +29,22 @@ fi
 filebase=/root/scripts
 
 echo "Docker pull dependency images, you can comment this if already have them ..."
-# c.f. https://github.com/sql-machine-learning/sqlflow/blob/develop/.travis.yml
-docker pull sqlflow/sqlflow:jupyter
-docker pull sqlflow/sqlflow:mysql
-docker pull sqlflow/sqlflow:server
-docker pull sqlflow/sqlflow:step
-docker pull argoproj/argoexec:v2.7.7
-docker pull argoproj/argocli:v2.7.7
-docker pull argoproj/workflow-controller:v2.7.7
+if [[ -d "/root/.cache" ]]; then
+    echo "Cache found at /root/.cache ..."
+    if [[ ! -f "/root/.cache/.loaded" ]]; then
+        find /root/.cache/* | xargs -I'{}' docker load -i '{}'
+        touch /root/.cache/.loaded
+    fi
+else
+    # c.f. https://github.com/sql-machine-learning/sqlflow/blob/develop/.travis.yml
+    docker pull sqlflow/sqlflow:jupyter
+    docker pull sqlflow/sqlflow:mysql
+    docker pull sqlflow/sqlflow:server
+    docker pull sqlflow/sqlflow:step
+    docker pull argoproj/argoexec:v2.7.7
+    docker pull argoproj/argocli:v2.7.7
+    docker pull argoproj/workflow-controller:v2.7.7
+fi
 echo "Done."
 
 # NOTE: According to https://stackoverflow.com/a/16619261/724872,
@@ -167,7 +175,7 @@ Congratulations, SQLFlow playground is up!
 Access Jupyter Notebook at: http://localhost:8888
 Access Kubernetes Dashboard at: http://localhost:9000
 Access Argo Dashboard at: http://localhost:9001
-Access SQLFlow with cli: ./sqlflow --datasource="\"$mysql_addr\""
+Access SQLFlow with cli: ./sqlflow --data-source="\"$mysql_addr\""
 
 Stop minikube with: minikube stop
 Stop vagrant vm with: vagrant halt