The best solution to extract 50 million records from a custom object everyday from Salesforce org without facing query timeout issues is to use an ETL tool for extraction of records. ETL stands for extract, transform, and load, and it refers to a process of moving data from one system to another. An ETL tool is a software application that can connect to various data sources, perform data transformations, and load data into a target destination. ETL tools can handle large volumes of data efficiently and reliably, and they often provide features such as scheduling, monitoring, error handling, and logging5. Using a custom auto number and formula field and use that to chunk records while extracting data is a possible workaround, but it requires creating additional fields and writing complex queries. The REST API can extract data as it automatically chunks records by 200, but it has some limitations, such as a maximum of 50 million records per query job6. Asking SF support to increase the query timeout value is not feasible because query timeout values are not configurable

Persistence is the data pattern that will allow a data architect to satisfy the requirements, while also keeping limits in mind. Persistence means storing data from external sources in Salesforce objects, either standard or custom. This allows you to access the data within Salesforce and use it for reporting, analytics, workflows, and other features. Persistence also helps you avoid hitting API limits or performance issues when accessing large volumes of data from external systems. You can use various tools such as Data Loader, Bulk API, or Platform Events to persist IoT data from connected devices to a cloud database in Salesforce.

Salesforce Connect is a feature that allows users to access data from external sources and multiple Salesforce orgs, using either clicks or code. Salesforce Connect can provide read-only access to customer records across all NTO orgs, without replicating or storing the data in Salesforce.