The Statistics page's once-a-minute refresh also ran on #/stats/explore and drew the Statistics page over it, losing the filters. It now runs on the Statistics page only, and the explore page refreshes itself every minute with the same filters (not while the filter bar is being edited).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Each explore filter takes several values, comma separated (any of them), !value to leave one out, and *.example.com for a name and the names below it. Every value is a chip with its own ✕; clicking a value in the lists adds it, and hide leaves it out. The filter bar's fields take lists. domain= still works.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The explore page has a filter bar to type filters in (client, name or it and below, view, type, result, blocked), starting from the current filters and suggesting values from what is shown, so taking every filter out no longer means going back. The Statistics page opens it with Explore the answers.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
jarld keeps every answer for a while (stats.retention.queries, default 7 days) in the statistics database: time, client, view, name, type, result and what blocked it. Every client, name, view, record type, result and blocklist on the Statistics page leads to an explore page for it: answers over time, the top values of every other dimension (each a link that narrows further) and the latest answers. GET /stats/explore.
The dashboard shows the size of Jarl's databases (and the query log's rows) and, per disk Jarl uses, how full it is and the databases' share.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>