On 23rd of August between 11:15 PM UTC and 1:00 AM the next day (1:45 h total) all our systems were often unreachable.
This was caused by a DNS outage that our upstream hosting provider Heroku faced on their entire platform. They have published a corresponding report on their status page at https://status.heroku.com/incidents/2453. Their underlying issue was in turn caused by their DNS provider NS1, who also have issued a status report over at https://ns1status.com/#!/incident/365716. The issues have been reported to be fully resolved.
We are sorry for any inconvenience this has caused