Describe the bug
Aggregations are killing performance in the web UI when the number of indexed torrents grows large (that's to say the total counts and faceted counts on the filters). They're a nice convenience but I don't think it's feasible to keep these in. This will involve some quite big UX changes - we'll probably keep a warm cache for the overall totals but you'll lose the counts on any filtered results. We'll also need to remove the total count from Torznab feeds edit: maybe later. The filtered aggregations will still be available in the GraphQL API.
To Reproduce
Index a few million torrents and see that filtering and search in the web UI and Torznab API is slow.
Describe the bug Aggregations are killing performance in the web UI when the number of indexed torrents grows large (that's to say the total counts and faceted counts on the filters). They're a nice convenience but I don't think it's feasible to keep these in. This will involve some quite big UX changes - we'll probably keep a warm cache for the overall totals but you'll lose the counts on any filtered results.
We'll also need to remove the total count from Torznab feedsedit: maybe later. The filtered aggregations will still be available in the GraphQL API.To Reproduce Index a few million torrents and see that filtering and search in the web UI and Torznab API is slow.