Mateusz Piasta
Mateusz has 16+ years of experience in QA and helps teams build high-load products. He believes in the power of knowledge sharing and is always looking to learn and grow alongside talented colleagues.
Many warehouses rely on warehouse management systems (WMS) to streamline operations. However, these systems can experience performance issues as businesses grow and data volume increases. This growth can lead to bottlenecks, slowdowns, and user frustration. So what happens when a custom WMS experiences performance issues? Mateusz Piasta, a QA Performance Expert at First Line Software, explores a project where we conducted performance testing on a custom WMS developed for one of our clients’, uncovering bottlenecks and paving the way for a smoother workflow.
For example, a CEO of a warehouse dreams of expanding his operation. With a humming warehouse powered by a trusty warehouse management system (WMS), you can envision a network of locations bustling with activity. A nagging worry lingers: can the WMS handle the explosion of data and user load that accompanies expanding warehouses? Disruptions during peak season could spell disaster. That’s where performance testing steps in.
Warehouse testing used to be a seasonal scramble. In the past, companies might only test their systems right before going live or during peak season. Imagine waiting until Black Friday to see if your store could handle the crowds! Testing involved manually simulating busy periods, which was expensive and limiting.
Today, testing is automated and always running. Automated performance testing tools allow companies to continuously analyze the performance of their systems, like having a fire drill year-round. This goes beyond just peak seasons – regular testing uncovers weaknesses throughout the entire supply chain system, before they cause problems.
The client’s existing WMS served their business needs well with a single warehouse, but its performance under increased data volume and user load was a complete unknown. With thousands of warehouse small goods feeding data into the system – could it handle the influx and generate reports efficiently?
Performance testing is a high-powered training session for your WMS. It throws simulated data punches from hundreds of warehouses, mimicking real-world scenarios. How will the WMS handle the barrage and generate reports quickly? Will order processing remain smooth even with a surge in activity? Performance testing answers these questions, pinpointing weaknesses before they become problems.
First Line Software used performance testing as a training scenario for the client’s WMS. This allowed our client to assess its ability to handle the workload and identify potential bottlenecks.
Potential performance testing reveals hidden weaknesses by pushing the WMS beyond its typical limits. Regular testing ensures that software updates and infrastructure upgrades don’t negatively impact WMS performance.
Our initial focus was developing a Proof of Concept (POC) to demonstrate the value of performance testing for our client. We understood that nonconventional testing methods would be required due to the system’s specific design:
Realistic Data Generation: Realistic test data that mimicked real-world warehouse operations was crucial. We created custom PowerShell scripts with parallel coding to effectively populate the database with intricate relationships between parts, customers, and warehouses, reflecting the complexities of the client’s system.
Database Adaptability: To ensure scalability and future-proofing, the scripts were designed to work seamlessly across both Microsoft SQL and Oracle database systems.
Phased Approach: Considering data size and approval requirements, we began with a POC. Ideally, a full-scale test leveraging a production database clone would provide the most comprehensive results.
We employed Grafana, a performance measurement tool, to assess the WMS’s behavior under realistic load. Custom scripts simulated real-world user activity, allowing us to pinpoint areas where the system struggled.
The success of this POC highlighted the potential for significant performance improvements in the client’s WMS. The next step was a full-scale performance test on a production database clone. This provided a more comprehensive picture of the system’s behavior under heavy load, allowing for targeted optimizations and a future-proof WMS solution.
We simulated busy warehouse activities with varying order volumes, user activity, and workloads. This helped us measure how fast the system responds (response time), how much work it can handle (throughput), and how efficiently it uses resources (resource utilization).
Our research and development efforts in performance testing for custom WMS are ongoing. We are continuously exploring new techniques and tools to further enhance the optimization process and empower warehouses to achieve peak efficiency.
Seasonal peaks and launches coming up? Identify and address system weaknesses now to avoid costly downtime. Contact us for a free consultation!
Mateusz has 16+ years of experience in QA and helps teams build high-load products. He believes in the power of knowledge sharing and is always looking to learn and grow alongside talented colleagues.
Whether you have a problem that needs solving or a great idea you’d like to explore, our team is always on hand to help you.