This article explains why phone numbers should be stored as VARCHAR(15~20) rather than INT/BIGINT, covering underlying principles, performance tests, internationalization, security, large‑scale optimization techniques, and a concrete best‑practice checklist for enterprise systems.
When designing a database for billions of phone numbers, using INT or BIGINT leads to overflow, data‑integrity loss, and inefficient queries, while VARCHAR(20) preserves formatting, supports international extensions, simplifies validation, and avoids common development pitfalls.
The article explains why using a VARCHAR(20) column to store billions of phone numbers is safer and more flexible than INT or BIGINT, covering range limits, data integrity, query convenience, interview expectations, and common pitfalls such as insufficient field length, charset issues, missing indexes, and lack of encryption.
The article examines an interview question about storing 2 billion phone numbers, explains why int cannot hold 11‑digit numbers, advocates using VARCHAR(20) for flexibility and data integrity, and outlines common pitfalls and best practices for database design.
When designing a schema for two‑billion phone numbers, use a VARCHAR(20) column with utf8mb4 Unicode, a unique index, and proper validation/encryption rather than a 32‑bit INT, because strings preserve leading zeros, international prefixes, extensions, and support business extensibility, fault tolerance, and future changes.
This article examines why storing 2 billion phone numbers using INT or BIGINT is problematic and argues for using VARCHAR(20) to preserve data integrity, support international formats, simplify queries, and avoid common pitfalls in database design and development.
