What are the performance optimization techniques in GORM?

GORM provides various performance optimization techniques to help developers improve database operation efficiency.

Query Optimization

1. Select Specific Fields

Only query needed fields to reduce data transfer:

go
// Not recommended: Query all fields
var users []User
db.Find(&users)

// Recommended: Only query needed fields
var users []User
db.Select("id", "name", "email").Find(&users)

2. Use Indexes

Create indexes for frequently used query conditions:

go
type User struct {
    gorm.Model
    Name  string `gorm:"index:idx_name"`
    Email string `gorm:"uniqueIndex"`
    Age   int    `gorm:"index:idx_age"`
}

Use Limit and Offset for pagination:

go
// Basic pagination
page := 1
pageSize := 10
offset := (page - 1) * pageSize

var users []User
db.Limit(pageSize).Offset(offset).Find(&users)

// Cursor pagination (more efficient)
var users []User
db.Where("id > ?", lastID).Limit(pageSize).Find(&users)

4. Avoid N+1 Queries

Use Preload to preload associated data:

go
// Not recommended: N+1 queries
var users []User
db.Find(&users)
for _, user := range users {
    var posts []Post
    db.Where("user_id = ?", user.ID).Find(&posts)
}

// Recommended: Use Preload
var users []User
db.Preload("Posts").Find(&users)

// Conditional preload
db.Preload("Posts", "status = ?", "published").Find(&users)

// Nested preload
db.Preload("Posts.Comments").Find(&users)

5. Use Pluck for Single Column

Use Pluck when only needing single column data:

go
// Not recommended
var users []User
db.Find(&users)
var names []string
for _, user := range users {
    names = append(names, user.Name)
}

// Recommended
var names []string
db.Model(&User{}).Pluck("name", &names)

Batch Operation Optimization

1. Batch Insert

Use CreateInBatches for batch insertion:

go
// Not recommended: Loop insertion
for _, user := range users {
    db.Create(&user)
}

// Recommended: Batch insert
db.CreateInBatches(users, 100)

2. Batch Update

Use batch update instead of loop update:

go
// Not recommended
for _, user := range users {
    db.Model(&user).Update("status", "active")
}

// Recommended
db.Model(&User{}).Where("id IN ?", userIDs).Update("status", "active")

3. Batch Delete

Use batch delete instead of loop delete:

go
// Not recommended
for _, user := range users {
    db.Delete(&user)
}

// Recommended
db.Where("id IN ?", userIDs).Delete(&User{})

Connection Pool Optimization

Configure Connection Pool

go
sqlDB, err := db.DB()
if err != nil {
    panic(err)
}

// Set maximum number of idle connections in the connection pool
sqlDB.SetMaxIdleConns(10)

// Set maximum number of open database connections
sqlDB.SetMaxOpenConns(100)

// Set maximum time a connection can be reused
sqlDB.SetConnMaxLifetime(time.Hour)

Query Caching

Use GORM Cache Plugin

go
// Use gorm-cache plugin
import "github.com/go-gorm/caches"

db.Use(caches.New(caches.Config{
    Redis:      redisClient,
    ExpireTime: 10 * time.Minute,
}))

Raw SQL Optimization

Use Raw SQL for Complex Queries

go
// Use raw SQL for complex queries
var results []struct {
    UserName string
    PostCount int
}
db.Raw(`
    SELECT u.name as user_name, COUNT(p.id) as post_count
    FROM users u
    LEFT JOIN posts p ON u.id = p.user_id
    WHERE u.age > ?
    GROUP BY u.id
    HAVING COUNT(p.id) > ?
`, 18, 5).Scan(&results)

Transaction Optimization

Reduce Transaction Scope

go
// Not recommended: Large transaction
tx := db.Begin()
// ... many operations ...
tx.Commit()

// Recommended: Small transaction
db.Transaction(func(tx *gorm.DB) error {
    // Only include necessary operations
    return nil
})

Database Design Optimization

1. Reasonable Use of Foreign Keys

go
type Order struct {
    gorm.Model
    UserID uint `gorm:"index"`
    User   User `gorm:"foreignKey:UserID;references:ID"`
}

2. Use Appropriate Data Types

go
type User struct {
    ID        uint   `gorm:"primaryKey"`
    Age       int8   `gorm:"type:tinyint"` // Use tinyint to save space
    Status    string `gorm:"type:char(1)"` // Fixed-length string
    CreatedAt time.Time
}

3. Partition Tables

For large tables, consider partitioning:

sql
-- MySQL partition example
CREATE TABLE orders (
    id BIGINT PRIMARY KEY,
    created_at DATETIME,
    -- other fields
) PARTITION BY RANGE (YEAR(created_at)) (
    PARTITION p2023 VALUES LESS THAN (2024),
    PARTITION p2024 VALUES LESS THAN (2025),
    PARTITION pmax VALUES LESS THAN MAXVALUE
);

Monitoring and Debugging

1. Enable Logging

go
// Enable detailed logging in development
db.Logger = logger.Default.LogMode(logger.Info)

// Only log slow queries in production
db.Logger = logger.Default.LogMode(logger.Silent)
db.Callback().Query().Before("gorm:query").Register("slow_query", func(db *gorm.DB) {
    start := time.Now()
    db.Statement.Callbacks().Query().After("gorm:query").Register("log_slow_query", func(db *gorm.DB) {
        if time.Since(start) > time.Second {
            log.Printf("Slow query: %s", db.Statement.SQL.String())
        }
    })
})

2. Use Explain to Analyze Queries

go
var users []User
result := db.Explain("SELECT * FROM users WHERE age > ?", 18)
fmt.Println(result)

Performance Testing

Benchmark Testing

go
func BenchmarkGORMQuery(b *testing.B) {
    db, _ := gorm.Open(mysql.Open(dsn), &gorm.Config{})
    
    b.ResetTimer()
    for i := 0; i < b.N; i++ {
        var users []User
        db.Select("id", "name").Limit(10).Find(&users)
    }
}

Best Practices Summary

Only query needed fields: Use Select to specify fields
Use indexes reasonably: Create indexes for frequently used query conditions
Avoid N+1 queries: Use Preload to preload associated data
Batch operations: Use batch insert, update, delete
Configure connection pool: Adjust connection pool parameters based on application load
Use pagination: Avoid loading large amounts of data at once
Optimize transactions: Keep transaction scope as small as possible
Monitor performance: Enable logging and slow query monitoring
Use raw SQL: Use raw SQL for complex queries
Regular maintenance: Regularly analyze tables and optimize indexes

Common Performance Issues

Q: How to solve N+1 query problem?

A: Use Preload to preload associated data, or use Joins for association queries.

Q: What to do if batch insert is slow?

A: Use CreateInBatches method and set reasonable batch size (usually 100-1000).

Q: How to optimize slow queries?

A: Use Explain to analyze query plan, check index usage, optimize query conditions.

Q: How large should the connection pool be?

A: Adjust based on application concurrency and database server performance, usually MaxOpenConns set to 2-4 times CPU cores.