These scripts and instructions are a starting point. They were originally written for launching Pentaho Enterprise Edition (EE) suite. They will not work out-of-the-box for Community Edition (CE). (For example, they require a directory containing the EE licenses.) In addition, the scripts depend on Pentaho Debian packages which are not publicly available. However, the Debian package Ant code is available in the trunk and you can use it to create your own packages. (See the package-deb target.) Finally, the location (pkg_url) where the Pentaho Debian packages are to be fetched must be changed in res/pentaho-init.

The technique used by these scripts for launching Pentaho on IaaS providers (e.g. Amazon aka AWS) is to use vanilla Ubuntu (tested with 11.04) along with the cloud-init package. Using this technique, all required software (e.g. Java, MySQL) is installed on first launch. As a consequence of this approach, first launch can take 10 minutes or more.

• Download Python 2.7.2 (http://www.python.org/getit/windows/) and run the installer
• Add folder with python.exe to PATH
• (Optional, AWS only) If launching instances on AWS, you need to install the boto library
  • Download boto 2.0 (http://code.google.com/p/boto/downloads/list)
  • Unzip the boto archive
  • Run: python setup.py install

• Run: sudo apt-get install python
• (Optional, AWS only) If launching instances on AWS, you need to install the boto library
  • Run: sudo apt-get install python-pip
  • Run: sudo pip install -U boto

Please see the Disclaimer first! There are two ways to use these scripts to run Pentaho on EC2.

The methods described in this document have been tested with Ubuntu 11.04. To find a list of Ubuntu 11.04 AMIs, following these instructions.

1. Go to http://cloud.ubuntu.com/ami/.
2. Filter using the following: Name=natty, EBS=ebs.

1. Run run_ec2.py. Use --help for usage.

1. Login to AWS Management Console and click the EC2 tab.
2. Create a key pair. A key pair will allow you to SSH into your instances.
   1. Click Key Pairs under Networking & Security.
   2. Click Create Key Pair.
   3. Enter a Key Pair Name and click Create.
   4. Save the downloaded pem file. This will be used every time you SSH.
   5. Click Close.
3. Create a security group. A security group will only allow certain traffic to your instances.
   1. Click Security Groups under Networking & Security.
   2. Click the checkbox for the default security group.
   3. Click the Inbound tab.
   4. For each of the following pairs, enter the Port range and Source then click Add Rule: (18080, 0.0.0.0/0), (19080, 0.0.0.0/0), (18088, 0.0.0.0/0), (18443, 0.0.0.0/0), (19443, 0.0.0.0/0).

1. Create the user data.

2. Run mk_ec2_userdata.py. Use --help for usage.

3. Login to AWS Management Console and click the EC2 tab. Now click Launch Instance.

4. Click Community AMIs. Enter one of the supported AMIs from above into the filter text box. Hit Enter. Click Select on the desired image.

5. Enter Number of Instances and Instance Type (tested with m1.large). Click Continue.

6. For User Data, click as file. Browse for the file created earlier. Do not check base64 encoded. Enable Termination Protection if desired. Shutdown Behavior should be set to Stop. Click Continue.

7. You can specify any number of tags. The Name tag is very useful when showing instances in the console. Click Continue.

8. Choose the key pair created earlier. Click Continue.

9. Choose the security group created earlier. Click Continue.

10. Confirm the settings and click Launch.

Please see the Disclaimer first!

This technique is based on Bootstrapping an Ubuntu Server on Rackspace Using Cloud-Init.

In the examples below, json.tool is used to pretty-print the JSON responses. Piping to it is optional.

1. Get an auth token.

   curl -s -D - -H "X-Auth-Key: @authKey@" -H "X-Auth-User: @authUser@" https://auth.api.rackspacecloud.com/v1.0

2. Export X-Auth-Token and X-Server-Management-Url values from the response.

   export X_AUTH_TOKEN="@authToken@" export X_MGMT_URL="@mgmtUrl@"

3. List the flavors. Choose a flavor (tested with 2GB of RAM). Note the flavor ID.

   curl -s -H "X-Auth-Token: $X_AUTH_TOKEN" $X_MGMT_URL/flavors/detail | python -mjson.tool

4. List the images. Choose an Ubuntu image (tested with 11.04). Note the image ID.

   curl -s -H "X-Auth-Token: $X_AUTH_TOKEN" $X_MGMT_URL/images | python -mjson.tool

5. Start an instance. Note the server ID, public IP address, and admin password.

   curl -s -H "X-Auth-Token: $X_AUTH_TOKEN" -H "Content-Type: application/json" -X POST -d '{"server":{"name":"pentaho","imageId":@imageId@,"flavorId":@flavorId@}}' $X_MGMT_URL/servers | python -mjson.tool

6. Check that the server status is ACTIVE.

   curl -s -H "X-Auth-Token: $X_AUTH_TOKEN" $X_MGMT_URL/servers/@serverId@ | python -mjson.tool

7. SSH into instance.

   ssh root@@serverIp@

8. Install cloud-init on the instance.

   apt-get install cloud-init

9. Create an image of running instance.

   curl -s -H "X-Auth-Token: $X_AUTH_TOKEN" -H "Content-Type: application/json" -X POST -d '{"image":{"serverId":@serverId@,"name":"pentaho"}}' $X_MGMT_URL/images | python -mjson.tool

10. Check that the image status is ACTIVE.

    curl -s -H "X-Auth-Token: $X_AUTH_TOKEN" $X_MGMT_URL/images/@imageId@ | python -mjson.tool

11. Delete server.

    curl -s -H "X-Auth-Token: $X_AUTH_TOKEN" -X DELETE $X_MGMT_URL/servers/@serverId@

1. Run mk_rs_request.py. Use --help for usage. Assume the output of that command goes to request.xml.

2. Launch the server.

   curl -s -H "X-Auth-Token: $X_AUTH_TOKEN" -H "Content-Type: application/xml" -X POST -d @request.xml $X_MGMT_URL/servers | python -mjson.tool