Google Cloud SQL is a managed database service that lets DBAs set up, maintain, and administer MySQL and PostgreSQL databases on Google Cloud Platform.

Delta Lake is an open source storage layer that sits on top of existing data lake file storage, such AWS S3, Azure Data Lake Storage, or HDFS. It uses versioned Apache Parquet files to store data, and a transaction log to keep track of commits, to provide capabilities like ACID transactions, data versioning, and audit history.

In most cases, the easiest way to retrieve data from relational databases is by writing SQL queries.

Google also provides a REST API for administering databases, instances, and other objects in Cloud SQL. So, for example, to retrieve a resource containing information about a database inside a Cloud SQL instance for a particular project, you could call GET /v1beta4/projects/[project]/instances/[instance]/databases/[database].

If your underlying database is PostgreSQL, you can use the pg_dump command to export data as a CSV-format flat file or a script that you can run to restore the database on any Postgres server. If your underlying database is MySQL, you can use the mysqldump command to export entire tables and databases in a format you specify (i.e. delimited text, CSV, or SQL queries that would restore the database).

The GET call we mentioned would return a database resource, which contains seven properties. Other API calls return different resources.

For data you export via SQL query, pg_dump, or mysqldump, you need a matching table in your data warehouse to receive the data from Cloud SQL. The information_schema database contains all of the metadata information you need to recreate your tables in another environment.

If you don't already have a data structure in which to store the data you retrieve, you'll have to create a schema for your data tables. Then, for each value in the response, you'll need to identify a predefined datatype (INTEGER, DATETIME, etc.) and build a table that can receive them. Google's documentation should tell you what fields are provided by each endpoint, along with their corresponding datatypes.

Complicating things is the fact that the records retrieved from the source may not always be "flat" – some of the objects may actually be lists. This means you'll likely have to create additional tables to capture the unpredictable cardinality in each record.

To create a Delta table, you can use existing Apache Spark SQL code and change the format from parquet, csv, or json to delta. Once you have a Delta table, you can write data into it using Apache Spark's Structured Streaming API. The Delta Lake transaction log guarantees exactly-once processing, even when there are other streams or batch queries running concurrently against the table. By default, streams run in append mode, which adds new records to the table. Databricks provides quickstart documentation that explains the whole process.

At this point you've coded up a script or written a program to get the data you want and successfully moved it into your data warehouse. But how will you load new or updated data? It's not a good idea to replicate all of your data each time you have updated records. That process would be painfully slow and resource-intensive.

Instead, identify key fields that your script can use to bookmark its progression through the data and use to pick up where it left off as it looks for updated data. Auto-incrementing fields such as updated_at or created_at work best for this. When you've built in this functionality, you can set up your script as a cron job or continuous loop to get new data as it appears in Google Cloud SQL.

And remember, as with any code, once you write it, you have to maintain it. If Google modifies its API, or the API sends a field with a datatype your code doesn't recognize, you may have to modify the script. If your users want slightly different information, you definitely will have to.

Delta Lake on Databricks is great, but sometimes you need to optimize for different things when you're choosing a data warehouse. Some folks choose to go with Amazon Redshift, Google BigQuery, PostgreSQL, or Snowflake, which are RDBMSes that use similar SQL syntax, or Panoply, which works with Redshift instances. Others choose a data lake, like Amazon S3. If you're interested in seeing the relevant steps for loading data into one of these platforms, check out To Redshift, To BigQuery, To Postgres, To Snowflake, To Panoply, and To S3.