Set up the Data Source¶
In this section, you will prepare the MinIO instance to use with Presto.
This section is comprised of the following steps:
1. Create a storage bucket¶
As we have already seen, MinIO will act as our underlying data storage for this workshop. MinIO is a high-performance, open-source object storage system. Because it implements the AWS s3 API (i.e., it is "s3-compatible"), we are able to use it to integrate with Presto. These types of object stores use buckets to organize files, giving us a specific location to provide to Presto when we read and write tables.
As part of our setup, a bucket called warehouse has already been created in our MinIO storage. See the mc service entrypoint in the docker-compose.yaml to for details on how this bucket has been created. Let's verify this by accessing the MinIO UI in a browser at the relevant IP address: http://<your_ip>:8443. If you're running everything on your local machine, the address will be http://localhost:8443.
Note
If you are using a remote Linux host, use the command curl https://ipinfo.io/ip to get your public IP address.
Note
Recall that we assigned a port mapping of 8443:9091 in docker-compose.yml, meaning that within the Docker network, the MinIO UI exists on port 9091, but it is exposed on port 8443 when accessing it from outside of the container.
You will be prompted for a username and password, which are minio and minio123 respectively, once again as defined in our docker-compose.yml for the minio service environment variables. Once you are logged in, you should see that the bucket titled warehouse is present.
If you see a webpage like the below that indicates that there are no available buckets, you can create one now. Click "Create a Bucket".
Enter the name warehouse and create the bucket. The name of the bucket must be warehouse in this case because that is the value that we passed to the REST server on container start for the catalog warehouse (CATALOG_WAREHOUSE=s3://warehouse/).
That's it! You can view the empty bucket in the "Object brower". Now our s3 object store is ready for use. Let's move to the next section to start creating Iceberg tables in Presto.
Note
If you run into errors in the third section of this workshop, there may have been more errors in your docker setup than just a missing bucket. Running docker restart minio, then docker restart mc a few moments later should resolve this issue.