A store that cannot handle traffic peaks loses money at the best possible moment. Black Friday, Cyber Monday, seasonal sales, sending a newsletter to 100,000 subscribers – these are moments when traffic in an hour can increase 10-fold. Preparing for peaks is not a matter of “if” but “when”.
How to check if the store is ready
A load test simulates many simultaneous users and shows at what number of sessions the store starts slowing down or stops responding. Tools like k6, Apache JMeter or LoadImpact let you simulate 100, 500, 1,000 concurrent users and measure response time, error count and server resource usage.
Without a load test you do not know at what traffic level the store will fail. With a test – you know before Black Friday, not during it.
Typical bottlenecks during traffic peaks
Database – SQL queries that take 0.1 seconds at normal traffic take 2 seconds at 10x traffic and block subsequent queries. Cache that is not correctly configured and passes traffic directly to PHP/database. Sessions stored in the database instead of Redis. External services (payments, couriers) that have their own rate limit per second.
What to do before the season
Review of Varnish and Redis configuration. Load test with a remediation plan for found bottlenecks. Monitoring review – will alerts fire before the store fails, not after. Establishing an escalation procedure – who gets a phone call at 2 AM on Black Friday if the store stops working. A customer communication plan in case of downtime.
If any of the signals described sound familiar, it is worth analysing before the problem grows. I offer a free initial consultation – write to biuro@tews.pl or visit tews.pl.
