We've been working on the Reimpact/platform project, and recently tackled a significant refactoring of our database schema. Our initial design had a mix of shared and module-specific tables within each tenant schema, which led to some challenges as the platform grew.

The primary goal of this refactoring was to improve modularity and maintainability by clearly separating module-specific data.

The Reimpact platform is undergoing a significant architectural shift, moving from a MySQL row-level multi-tenancy model to a more robust PostgreSQL schema-per-tenant setup. This post dives into the key considerations and steps involved in this migration, leveraging Laravel and Filament.

The Motivation for Schema-Based Tenancy

Row-level multi-tenancy in MySQL, while initially simpler to

The platform project is undergoing a series of updates to enhance its architecture and developer experience.

Streamlining Panel Providers

Component registration is now more organized. The PackagingPanelProvider has been registered, ensuring that necessary components are readily available throughout the application. This centralizes the management of UI elements, improving maintainability.

Migrating data between different database systems can be a complex undertaking. Recently, the team working on the platform navigated a significant data migration from MySQL to PostgreSQL. This post outlines some key considerations and challenges encountered during the process. These insights can help you avoid common pitfalls when undertaking similar migrations.

The Challenge

Introduction

When deploying the Reimpact platform, ensuring a smooth setup process across different environments is crucial. This post addresses a common issue encountered during tenant schema population, specifically when the executing user lacks PostgreSQL superuser privileges. We'll explore the problem and how a simple try-catch block resolves it, enhancing the platform's robustness.

Introduction

When building multi-tenant applications with Laravel, managing database schemas for each tenant can become complex. We recently improved our tenant schema population process in the platform project to handle various edge cases and ensure data integrity across companies.

The Problem

Our initial schema population process had several limitations:

1. PostgreSQL's

Introduction

Row-level filtering is common for multi-tenancy, but what if you need stronger isolation? At Reimpact's platform, we're migrating towards PostgreSQL schemas to achieve true tenant separation, leveraging Laravel, Filament, and the middleware pattern.

This post details our approach to dynamically setting the PostgreSQL search_path for each tenant, ensuring data isolation and

Introduction

When building multi-tenant applications, choosing the right architecture is critical. This post explores how to leverage PostgreSQL schemas within a Laravel application to achieve tenant isolation, enhance security, and improve scalability.

The Multi-Tenant Challenge

Traditional multi-tenant approaches often involve sharing a single database and distinguishing tenants by a

Introduction

When developing dashboards that rely on precise data comparisons in PostgreSQL, case sensitivity can be a silent but critical issue. This post examines a case in the Reimpact platform where case sensitivity in unit comparisons caused dashboard calculations to return incorrect results, and how it was resolved.

The Problem: Case-Sensitive Comparisons

PostgreSQL, by default,

When developing applications using Laravel with a PostgreSQL database, maintaining strict data type consistency is crucial, especially when dealing with dashboard functionalities that aggregate data. PostgreSQL's strong type checking can lead to unexpected errors if data types don't match exactly between the database schema and the application's expectations.

The Problem: Type Mismatches