19:59InternetCZ · Tor users in Czechia fell below the expected range (44,012 vs at least 68,661)backfillTor Metrics
19:59InternetLT · Tor users in Lithuania jumped past the expected range (31,990 vs up to 17,178)backfillTor Metrics
Fri 11 Sep 2026
19:59InternetSG · Tor users in Singapore jumped past the expected range (54,624 vs up to 27,670)backfillTor Metrics
Tue 8 Sep 2026
19:59InternetCZ · Tor users in Czechia jumped past the expected range (96,153 vs up to 54,982)backfillTor Metrics
What this page is not. It does not predict anything and it does not beat the news: almost every record here is published openly and reaches reporters the same minute it reaches this page. Where the slow signals come from. Internet outages, military aviation, emptied airspace, GPS interference and ship traffic are measured on our own collector server and publish only after each has learned its normal. Alerts by email, Telegram or webhook, and the records through an API and MCP, are on the Geo plans. Rows marked backfill came from a source's own history rather than from live collection, so they never count toward how fast we saw something.
A source that stops answering does not go quiet here. Its rows stop and this list says when it was last heard from, because no data is not the same as nothing happening.
Times are New York (ET), like every page on this site. US federal records are public domain, UK advice is Open Government Licence, GDACS and Wikimedia carry Creative Commons terms, conflict deaths are from the Uppsala Conflict Data Program (UCDP GED Candidate, CC BY 4.0). Every row links to the record it came from and nothing is republished as a bulk feed. This week as JSON (times in UTC, each item marked routine or not; every week)