1. Separate request volume from cost
Several expensive category URLs can matter more than many cached visits. Record when the slowdown happened and a few example filter, sorting or pagination URLs.
Compare the same category with and without ordinary filters, including a new session and a repeat visit. This is a baseline check, not a load test. Do not generate a request flood against your live store.
2. Collect a focused log sample
Ask your hosting provider or administrator for the busiest routes during the slowdown. Look for changing parameter combinations, deep pagination, unusual page sizes and repeated request sequences. Parameter names depend on the installed filtering module.
Do not classify every crawler as an attack. A User-Agent name does not verify identity, and an IP address does not describe request cost. Compare routes, frequency, response status and server measurements.
- Use the same time window for request logs and hosting charts.
- Compare automated traffic with customer activity and scheduled imports.
- Remove customer data, tokens and session parameters before sharing logs.
3. Separate SEO decisions from abuse protection
Agree with your SEO owner which filtered views deserve to be landing pages. Review crawl management for the remaining combinations. Google's faceted navigation guidance explains how filter combinations can create a very large URL space.
robots.txt depends on a crawler respecting its rules. Canonical is a preferred-URL signal, not a request limiter. Do not copy a blanket rule blocking every parameter: needed pages and functions may also depend on parameters.
4. Make one change at a time
With Traffic Guard Pro, observe the layer under review and check whether other mechanisms already block traffic. Use route-cost diagnostics and compare detections with real customer journeys before choosing protection scope.
Record the time, previous settings and rollback procedure. Test categories, filters, products and cart, including on mobile, plus required integrations. A higher block count alone does not demonstrate an improvement.
5. Compare like with like
Compare response times, server load and successful journeys at similar hours and traffic levels. If filters remain slow without automated traffic, investigate the filtering module, database queries, cache and hosting.
Traffic protection cannot solve every slow-category problem. A useful support report includes PrestaShop, PHP and module versions, an example route, timestamp and before/after observations.