Optimize & Extend PHP ERP
Budget: $250 – $750 USD
Our in-house PHP-based ERP has grown fast, and two areas now limit us: reporting and the user-permission layer. Both modules place a heavy load on the server, so I need them re-engineered to cut resource usage and scale smoothly as concurrent traffic rises.
The work focuses on:
• Reporting – streamline queries, introduce smart caching or pagination, and keep every existing report layout intact.
• User permissions – redesign the logic so role checks are quick, minimise round-trips to MySQL, and make the structure easier to expand later.
Performance goals:
• Lower average CPU/RAM consumption per request.
• Maintain sub-second response times during stress tests with double our current user load.
Deliverables
1. Refactored PHP code (PHP 8.1, PSR-12) committed via Git.
2. Any SQL migration scripts needed for schema tweaks.
3. A concise README/markdown explaining changes, before-vs-after benchmarks, and how to reproduce the load test (ApacheBench, JMeter, etc.).
4. Automated tests proving existing business logic still passes.
Acceptance criteria
• ≥30 % reduction in memory footprint under normal use.
• Stable operation at 100 concurrent sessions for 15 minutes, average response ≤1 s.
• No regression in visible functionality.
I’m happy to clarify architecture or share sample data once the project begins, and I’ll review your pull request as soon as benchmarks hit the targets.
The work focuses on:
• Reporting – streamline queries, introduce smart caching or pagination, and keep every existing report layout intact.
• User permissions – redesign the logic so role checks are quick, minimise round-trips to MySQL, and make the structure easier to expand later.
Performance goals:
• Lower average CPU/RAM consumption per request.
• Maintain sub-second response times during stress tests with double our current user load.
Deliverables
1. Refactored PHP code (PHP 8.1, PSR-12) committed via Git.
2. Any SQL migration scripts needed for schema tweaks.
3. A concise README/markdown explaining changes, before-vs-after benchmarks, and how to reproduce the load test (ApacheBench, JMeter, etc.).
4. Automated tests proving existing business logic still passes.
Acceptance criteria
• ≥30 % reduction in memory footprint under normal use.
• Stable operation at 100 concurrent sessions for 15 minutes, average response ≤1 s.
• No regression in visible functionality.
I’m happy to clarify architecture or share sample data once the project begins, and I’ll review your pull request as soon as benchmarks hit the targets.