To retrieve data from Google PageSpeed Insights, you need to access its API. First, ensure you have a Google Cloud account and create a project. Enable the PageSpeed Insights API in your Google Cloud console. Obtain an API key that will allow you to authenticate your requests to the API.

Use a programming language like Python to send HTTP GET requests to the PageSpeed Insights API. Construct your request URL with the required parameters, including the URL of the page you want to analyze and your API key. Parse the JSON response to extract the performance metrics and insights you need.

Once you have the JSON response, transform the data into a structured format suitable for Iceberg. This typically involves converting the JSON data into a tabular format. You can use libraries such as Pandas in Python to transform and clean the data as needed, ensuring consistency and readiness for loading into Iceberg.

Set up an Apache Iceberg environment on your local machine or a server. This will involve installing Apache Iceberg and configuring it with a compatible compute engine like Apache Spark. Ensure your environment is ready to receive and store data in a tabular format.

Apache Iceberg works efficiently with columnar storage formats like Apache Parquet. Convert your structured data from the previous step into Parquet format. This can be accomplished using libraries such as PyArrow or directly within a Spark job if using Apache Spark. Parquet provides efficient data compression and encoding, which are beneficial for storage and query performance.

With your data in Parquet format, you can now load it into an Iceberg table. Define the schema for your Iceberg table to match the structure of your Parquet files. Use Apache Spark to create or append to an Iceberg table by reading the Parquet files and writing them into the Iceberg format. This can be done using Spark's DataFrame API and its integration with Iceberg.

After loading the data, verify that it has been accurately stored in Iceberg. Use SQL queries to check the data integrity and ensure it matches the original metrics from the PageSpeed Insights API. Apache Iceberg allows you to perform complex queries on your data, leveraging its support for schema evolution and partitioning for efficient querying.

By following these steps, you will be able to manually move data from Google PageSpeed Insights to Apache Iceberg without relying on third-party connectors or integrations.