Interested in launching Hadoop with Amazon Web Services (AWS)? We've expanded cloud deployment options for MapR on AWS.
MapR worked closely with AWS to develop AMIs (Amazon Machine Images) that enable hourly usage of the MapR Distribution on AWS. These AMIs, which are pre-loaded and pre-configured with the MapR software and the required supporting operating system, should be launched via instructions on this page using Cloud Formation Templates. Cloud Formation Templates automate the provisioning of the resources required to form a MapR cluster, and ensure that the MapR software is installed properly.
Note: There are other ways you can deploy MapR on your AWS instance. See the MapR-AWS partner page for more information.Instructions for launching MapR on AWS
Please note: AWS EC2 instances are charged per hour or annually. Please make sure to follow instructions to turn off your EC2 instance server running MapR after use. If you do not deactivate your instances, you will continue to be charged by the hour for EC2 usage.
When you stop your EC2 instances, data that was stored in local storage on each instance WILL BE LOST and MapR will not be accessible until you restart the EC2 instance servers. Data that was stored on EBS storage will still be available after stopping the EC2 instances, and as such, you will still be charged for EBS usage.Prerequisites
Amazon AWS Account. Don't have an AWS account? Sign up here.
You can now refer to the documentation below that will explain more of what you will see on AWS.
- This will bring you to the Cloud Formation Template selection screen (see figure below). If desired, rename your cluster from the default name that has automatically been input for you. Note that cluster names must be unique, so never reuse the same name. Click Next when you are ready.
- Complete the required information on my Cloud formation template.
- ClusterInstanceCount – the number of nodes that you want in the cluster. You will be charged on an hourly basis for the hardware and software on each node.
- ClusterID – unique name for your MapR cluster, e.g. Company-Prod
- Instance Spot Price – The clearance price is you want to use spot instead of reserved instance prices. More information on instance pricing is available here.
- KeyName – name of an existing EC2 key pair. Key pairs help keep your cluster secure.
- Persistent Storage – the amount of persistent storage that you want to allocate to each node, in gigabytes. Use this to prevent accidental data loss when clusters are shut down.
- RemoteAccessCIDR – use this to restrict SSH and web access to the cluster to a specific block of IP addresses. Leave blank for open access.
- MapR Edition – pre-populated field describing the MapR version you wish to launch. Double check that you are launching the version that matches the listing to which you subscribed. If the versions do not match, the cloud formation process will fail and you will need to start over.
- Click Next to start provisioning the required infrastructure and installing MapR.
To access the cluster, you will need to use SSH (Secure Shell) to open a connection to the cluster. If you are using a Windows machine, you may need to download software to use ssh, such as PuTTY. If you are using a Mac or Linux, ssh should be available from the command line as follows (Windows users will need to enter this information into their client of choice).
Select outputs from the tabs below the EC2 console to view the Cloud Formation Template outputs which contain the information required to access the cluster. You can connect directly into the cluster node at the host IP (for the example below 22.214.171.124) using the MCS credentials (user: mapr, password: i-6b358182). All the cluster nodes are configured similarly; the password for the mapr user is the Amazon instance-id of the node. Windows users should use putty or other ssh-capable terminal program when connecting to the cluster nodes.