Optimizing Laravel Database Seeders for Performance

Seeding large datasets in Laravel can be really slow. Here's how to optimize your seeders for better performance.

Problem

Consider this typical seeder approach:

public function run()
{
    foreach ($teams as $team) {
        Project::factory()->count(100)->create([
            'team_id' => $team->id
        ]);
    }
}

This seems simple, but it's incredibly inefficient because it:

• Creates individual database transactions for each record
• Triggers model events for each creation
• Runs validations and other overhead for each record

Solution

Here are key strategies to dramatically improve seeder performance:

1. Use Bulk Insertions

Instead of creating records one by one, batch them:

// Bad: Individual insertions
Project::factory()->count(100)->create();

// Good: Bulk insertion
$projects = Project::factory()->count(100)->make()->toArray();
Project::insert($projects);

2. Chunk Large Datasets

When dealing with thousands of records, chunk them to manage memory:

collect($projects)->chunk(100)->each(function ($chunk) {
    Project::insert($chunk->toArray());
});

3. Prepare Data in Memory First

Collect all data before touching the database:

// Prepare all data first
$projects = [];
$projects = array_merge($projects, Project::factory()->count(100)->make()->toArray());
$projects = array_merge($projects, Project::factory()->count(50)->make()->toArray());

// Then insert in one go
collect($projects)->chunk(100)->each(function ($chunk) {
    Project::insert($chunk->toArray());
});

4. Handle JSON and DateTime Fields Properly

When using bulk insertions, you need to manually format special fields:

$projects = collect($projects)->map(function ($project) {
    return array_merge($project, [
        'created_at' => now()->format('Y-m-d H:i:s'),
        'settings' => isset($project['settings']) ? json_encode($project['settings']) : null,
    ]);
});

Real-World Example

Here's a complete example that implements all these optimizations:

public function run(): void
{
    $teams->each(function ($team) {
        $now = now()->format('Y-m-d H:i:s');
        $projects = [];

        // Prepare data in memory
        $projects = array_merge($projects, 
            Project::factory()->count(300)
                ->for($team)
                ->make()
                ->map(function ($item) use ($now) {
                    return [
                        'team_id' => $item->team_id,
                        'name' => $item->name,
                        'description' => $item->description,
                        'settings' => isset($item->settings) ? json_encode($item->settings) : null,
                        'created_at' => $now,
                        'updated_at' => $now
                    ];
                })
                ->toArray()
        );

        // Bulk insert in chunks
        collect($projects)->chunk(100)->each(function ($chunk) {
            Project::insert($chunk->toArray());
        });
    });
}

Trade-offs to Consider

These optimizations can make your seeders run 10-100x faster, but there are some trade-offs:

• Model events won't fire
• Timestamps aren't automatically set
• You'll need to handle relationships manually
• Validation rules aren't checked

What I'm building

Delegate tasks. Get software.

Give Vroni a GitHub issue, bug report, spec, or rough idea. It reads the repo, plans the change, writes code, runs checks, and works toward a review-ready pull request.

Take a look at vroni.com