Rendering of the Google Maps Tiles on the Amazon EC2 Cluster

- You have several input raster files which you would like to merge (NOAA RNCs, satellite/aerial imagery, scanned maps)
- You have Amazon AWS account
- You would like to save the tiles into S3 hosting

1) Upload the input data into an shared S3 bucket ("inputBucket") - all nodes in the cluster will use it as read-only disk (reading from such disk is for free for EC2 nodes). You could use a project such as s3fs. This storage should be later on mounted to all computing nodes to /input.

2) Start computing cluster nodes from prepared MapTile Cluster AMI (id-...). All nodes must have assigned IP, and open port 60000 in the firewall for ppserver. Everything else is started automatically.

3) Either from your own computer or from one of the EC2 nodes you can start the rendering with the master node.
Parameters are available after --help.
...
The rendering progress is printed on the computer where you started the master node.
Comments