---
title: Schema and typing
meta:
  description: Common issues and solutions for schema design and data typing in ClickHouse and Tinybird.
---

# Schema and typing

Common issues and solutions for schema design and data typing in ClickHouse and Tinybird.

## Overview

This section covers troubleshooting for schema design issues and data type problems that can cause errors and performance issues.

## Schema categories

### Type mismatch

Common type mismatch issues:

- **TYPE_MISMATCH** - Incompatible data types in operations
- **NO_COMMON_TYPE** - No common type for operations
- **String to numeric conversion** - Converting between types safely
- **Date type mismatches** - Mixing different date types

[View type mismatch troubleshooting →](/sql-reference/schema/type-mismatch)

### Nullable vs not nullable

Deciding when to use nullable columns:

- **When to use nullable** - Optional data and external sources
- **When to use not nullable** - Required fields and performance-critical columns
- **Performance considerations** - Impact of nullable columns
- **Migration strategies** - Converting between nullable types

[View nullable vs not nullable troubleshooting →](/sql-reference/schema/nullable-vs-not-nullable)

### Inferred wrong type

Issues with ClickHouse type inference:

- **String inferred as numeric** - Type inference problems
- **Numeric inferred as string** - Mixed data type issues
- **Schema hints** - Using SCHEMA to override inference
- **Debugging type inference** - Using toTypeName() function

[View inferred wrong type troubleshooting →](/sql-reference/schema/inferred-wrong-type)

### How to debug types

Tools and techniques for debugging types:

- **Using toTypeName()** - Checking column and expression types
- **Schema debugging** - Examining table schemas
- **Data validation** - Checking for type inconsistencies
- **Type conversion debugging** - Understanding conversion issues

[View how to debug types →](/sql-reference/schema/how-to-debug-types)

## Common patterns

### Type conversion issues

Strategies for handling type conversions:

1. **Use safe conversions** - Use `toTypeOrNull()` functions
2. **Check data types** - Use `toTypeName()` to verify types
3. **Handle nulls explicitly** - Check for null values before conversions
4. **Validate inputs** - Ensure data meets type requirements

### Schema design principles

Best practices for schema design:

1. **Use explicit types** - Don't rely on type inference
2. **Be consistent** - Use same types across related tables
3. **Consider performance** - Choose types based on usage patterns
4. **Plan for growth** - Design schemas that can evolve
5. **Document decisions** - Keep track of type choices and reasons

## Best practices

1. **Always specify types** - Use explicit type declarations
2. **Use safe conversions** - Prefer functions that handle errors gracefully
3. **Test with real data** - Verify types with actual data samples
4. **Monitor type issues** - Track type-related errors and performance
5. **Document type decisions** - Keep track of type choices and trade-offs

## Related documentation

- [Data types](/sql-reference/data-types)
- [Type conversion functions](/sql-reference/functions/type-conversion-functions)
- [ClickHouse errors](/sql-reference/clickhouse-errors)