Hi Sigrid
Thanks for letting me know what's discussed. Although I do not understand German, I translated some of the chat responses.
FYI. we are working on a few improvements for the next release. Among others:
1. We optimized the queries for nested filters (i.e. categories and tags).
2. We are introducing filter caching. At this stage, it will work only for the root filters (since they do not change). But in a later stage, we can explore further integration.
The solution about the POST/Redirect/GET cannot be applied here. It is about protecting forms from re-submission, which is not the case with JFilters.
Server outages and peaks were reported by other users as well.
The main cause, were bots crawling the many pages created by JFilters based on the filters combinations.
Most bots respect the canonical tags and the nofollow, hence:
1. Make sure that you have a root filter in a published or listening state.
2. Set Follow Links by Search Engines > No, either for each filter individually or in the component configuration for all filters (given that each filter has this setting set to Global).
3. Track from where the majority of such a traffic comes. Usually a bot that does not respect "nofollow" and canonical tags, is not a welcome one. Also a bot that does hundreds or thousands of requests in a small amount of time is not a benevolent one.
You can use your website's robots.txt to only allow legit bots (see: https://developers.netlify.com/guides/blocking-ai-bots-and-controlling-crawlers/)
4. If you can put your website behind a CDN like Cloudflaire, most of the unwanted traffic will be filtered before reaching your server.
I am using it for years will great results.
5. As a last resource measure, you can set your JFilters menu item/page to "noindex, nofollow" from its robots settings, but this will have an impact in your SEO, if those pages have value in terms of SEO.
The post became too big, I would rather write a blog post on the matter, at some point.
Kind regards,
Sakis Terzis
Founder and Lead Developer
