Phew, it's back

Good news, it looks like the Wikipedia outage was indeed short-lived — or at least only lasted an hour.

The Downdetector reports have plummeted back down to zero, which suggests that 'emergency maintenance' is now finished.

Fortunately, that was all a lot more straightforward than the major PHP flaw that's turned government websites dark and sent website admins scambling...

'Emergency maintenance'

This probably explains today's Wikipedia issues — the Wikimedia Foundation status page says 'emergency maintenance' is in progress.

The foundation hasn't explained how long the maintenance will last or what it's for, but that should hopefully mean it's pretty short-lived...

Please try again later...

This is what Wikipedia and Wikimedia currently look like for many users — an "error" page with a "please try again later" message. I've also never seen it go down like that, although it's working fine for me in the UK.

Strangely, there are a few other outages running concurrently right now, including Asana and the Internet Archive. It isn't yet clear if they're related, but they did start at very similar times (around 10am ET / 3pm BST). Wikipedia is certainly the worst-hit, though.

A sharp spike in reports

Wikipedia is working okay for me here in the UK, but it isn't for thousands of users right now — the reports graph on Downdetector is now over 3,000 in the US, and it's climbing in the UK too.

The cause is so far unclear, as there are no current issues at the major cloud providers like AWS, Google Cloud or Cloudflare. I'll be digging into this and reporting back shortly...